path: root/Docs/manual.texi

\input texinfo  @c -*-texinfo-*-
@c Copyright 1997-2002 TcX AB, Detron HB and MySQL Finland AB
@c
@c This manual is NOT distributed under a GPL style license.
@c Use of the manual is subject to the following terms:
@c - Conversion to other formats is allowed, but the actual
@c   content may not be altered or edited in any way.
@c - You may create a printed copy for your own personal use.
@c - For all other uses, such as selling printed copies or
@c   using (parts of) the manual in another publication,
@c   prior written agreement from MySQL AB is required.
@c
@c Please e-mail docs@mysql.com for more information or if
@c you are interested in doing a translation.
@c
@c *********************************************************
@c Note that @node names are used on our web site.
@c So do not change node names without checking
@c Makefile.am and SitePages first.
@c *********************************************************
@c
@c %**start of header

@c there's a better way to do this... i just don't know it yet
@c sed will remove the "@c ifnusphere " to make this valid
@c ifnusphere @set nusphere 1

@setfilename mysql.info

@c We want the types in the same index
@syncodeindex tp fn

@c Get version information. This file is generated by the Makefile!!
@include include.texi

@ifclear tex-debug
@c This removes the black squares in the right margin
@finalout
@end ifclear

@c Set background for HTML
@set _body_tags BGCOLOR=silver TEXT=#000000 LINK=#101090 VLINK=#7030B0
@c Set some style elements for the manual in HTML form. 'suggested'
@c natural language colors: aqua, black, blue, fuchsia, gray, green,
@c lime, maroon, navy, olive, purple, red, silver, teal, white, and
@c yellow. From Steeve Buehler <ahr@YogElements.com>
@set _extra_head <style> code {color:purple} tt {color:green} samp {color:navy} pre {color:maroon} </style>

@settitle MySQL Reference Manual for version @value{mysql_version}.

@c We want single-sided heading format, with chapters on new pages. To
@c get double-sided format change 'on' below to 'odd'
@ifclear nusphere
@setchapternewpage on
@end ifclear

@ifset nusphere
@setchapternewpage odd
@end ifset

@paragraphindent 0

@ifset nusphere
@smallbook
@end ifset

@c %**end of header

@ifinfo
@format
START-INFO-DIR-ENTRY
* mysql: (mysql).               MySQL documentation.
END-INFO-DIR-ENTRY
@end format
@end ifinfo

@titlepage
@sp 10
@center @titlefont{MySQL Reference Manual}
@sp 10
@center Copyright @copyright{} 1997-2002 MySQL AB
@c blank page after title page makes page 1 be a page front.
@c also makes the back of the title page blank.
@page
@end titlepage

@c Short contents, blank page, long contents.
@c until i can figure out the blank page, no short contents.
@c @shortcontents
@c @page
@c @page
@contents

@c This should be added. The HTML conversion also needs a MySQL version
@c number somewhere.

@iftex
@c change this to double if you want formatting for double-sided
@c printing
@headings single

@oddheading @thischapter @| @| @thispage
@evenheading @thispage @| @| MySQL Technical Reference for Version @value{mysql_version}

@end iftex


@node Top, Introduction, (dir), (dir)

@ifinfo
This is a manual for the @code{MySQL Database System}.
This version is about the @value{mysql_version} version of
@code{MySQL Server}. You can find a specific manual for any older
version of @code{MySQL Server} in the binary or source distribution
for that version.
@end ifinfo

@menu
* Introduction::                General Information
* Installing::                  Installing MySQL
* Tutorial::                    MySQL Tutorial
* MySQL Database Administration::  Database Administration
* MySQL Optimisation::          MySQL Optimisation
* Reference::                   MySQL Language Reference
* Table types::                 MySQL Table Types
* Clients::                     MySQL Client Tools and APIs
* Extending MySQL::             Extending MySQL
* Problems::                    Problems
* Contrib::                     Contributed Programs
* Credits::                     Contributors to MySQL
* News::                        MySQL Change History
* Porting::                     Porting to Other Systems
* Environment variables::       MySQL environment variables
* Regexp::                      MySQL Regular Expression Syntax
* GPL license::                 GNU General Public License
* LGPL license::                GNU Lesser General Public License
* Function Index::              SQL Command, Type and Function Index
* Concept Index::               Concept Index
@end menu


@node Introduction, Installing, Top, Top
@chapter General Information

@cindex overview
@cindex general information

The @code{MySQL (TM)} software delivers a very fast, multi-threaded,
multi-user, and robust @code{SQL} (@code{Structured Query Language})
database server.
@code{MySQL Server} is intended for mission-critical, heavy-load
production systems as well as for embedding into mass-deployed software.
@code{MySQL} is a trademark of @code{MySQL AB}.

The @code{MySQL} software has @code{Dual Licensing}, which means you can use
the @code{MySQL} software free of charge under the
@code{GNU General Public License} (@uref{http://www.gnu.org/licenses/}).
You can also purchase commercial @code{MySQL} licenses from
@code{MySQL AB} if you do not wish to be bound by the terms of the
@code{GPL}.
@xref{Licensing and Support}.

The @code{MySQL} web site (@uref{http://www.mysql.com/}) provides the
latest information about the @code{MySQL} software.

The following list describes some sections of particular interest in this manual:

@itemize @bullet
@item
For information about the company behind the @code{MySQL Database Server},
see @ref{What is MySQL AB}.

@item
For a discussion about the capabilities of the @code{MySQL Database Server},
see @ref{Features}.

@item
For installation instructions, see @ref{Installing}.

@item
For tips on porting the @code{MySQL Database Software} to new architectures
or operating systems, see @ref{Porting}.

@item
For information about upgrading from a Version 3.23 release, see
@ref{Upgrading-from-3.23}.

@item
For information about upgrading from a Version 3.22 release, see
@ref{Upgrading-from-3.22}.

@item
For a tutorial introduction to the @code{MySQL Database Server},
see @ref{Tutorial}.

@item
For examples of @code{SQL} and benchmarking information, see the
benchmarking directory (@file{sql-bench} in the distribution).

@item
For a history of new features and bug fixes, see @ref{News}.

@item
For a list of currently known bugs and misfeatures, see @ref{Bugs}.

@item
For future plans, see @ref{TODO}.

@item
For a list of all the contributors to this project, see @ref{Credits}.
@end itemize

@strong{Important}:

Reports of errors (often called bugs), as well as questions and comments,
should be sent to the mailing list at @email{mysql@@lists.mysql.com}.
@xref{Bug reports}.
The @code{mysqlbug} script should be used to generate bug reports.
@c FIX! RPMs are also binary
For source distributions, the @code{mysqlbug} script can be found in the
@file{scripts} directory.  For binary distributions, @code{mysqlbug} can
be found in the @file{bin} directory. If you have found a sensitive
security bug in @code{MySQL Server}, you should send an e-mail to
@email{security@@mysql.com}.

@cindex errors, reporting
@cindex reporting, errors
@cindex mailing list address
@cindex @code{mysqlbug} script, location


@menu
* Manual-info::                 About This Manual
* What-is::                     What Is MySQL?
* What is MySQL AB::            What Is MySQL AB?
* Licensing and Support::       MySQL Support and Licensing
* MySQL 4.0 In A Nutshell::     MySQL 4.0 In A Nutshell
* MySQL Information Sources::   MySQL Information Sources
* Compatibility::               How Standards-compatible Is MySQL?
* TODO::                        MySQL and the future (The TODO)
* Comparisons::                 How MySQL Compares to Other @code{Open Source} Databases
@end menu


@node Manual-info, What-is, Introduction, Introduction
@section About This Manual

@cindex online location of manual
@cindex manual, online location

This is the @code{MySQL} reference manual; it documents @code{MySQL}
Version @value{mysql_version}. Being a reference manual, it does not
provide general instruction on @code{SQL} or relational database
concepts.

As the @code{MySQL Database Software} is under constant development,
the manual is also updated frequently.
The most recent version of this manual is available at
@uref{http://www.mysql.com/documentation/} in many different formats,
including Texinfo, plain text, Info, HTML, PostScript, PDF,
and Windows HLP versions.

The primary document is the Texinfo file.
The HTML version is produced automatically using a modified version of
@code{texi2html}.
The plain text and Info versions are produced with @code{makeinfo}.
The PostScript version is produced using @code{texi2dvi} and @code{dvips}.
The PDF version is produced with @code{pdftex}.

If you have a hard time finding information in the manual, you can try
our searchable PHP version at @uref{http://www.mysql.com/doc/}.

If you have any suggestions concerning additions or corrections to this
manual, please send them to the documentation team at @email{docs@@mysql.com}.

@cindex manual, available formats
@cindex Texinfo

This manual is written and maintained by David Axmark, Michael (Monty)
Widenius, Jeremy Cole, Arjen Lentz, and Paul DuBois. For other
contributors, see @ref{Credits}.

The copyright (2002) to this manual is owned by the Swedish company
@code{MySQL AB}. @xref{Copyright}.

@menu
* Manual conventions::          Conventions Used in This Manual
@end menu

@node Manual conventions,  , Manual-info, Manual-info
@subsection Conventions Used in This Manual

@c We use British English. But let's not publicise that here yet ;-)

This manual uses certain typographical conventions:

@cindex manual, typographical conventions
@cindex typographical conventions
@cindex conventions, typographical

@table @asis
@item @code{constant}
Constant-width font is used for command names and options; SQL statements;
database, table, and column names; C and Perl code; and environment variables.
Example: ``To see how @code{mysqladmin} works, invoke it with the
@code{--help} option.''

@item @file{filename}
Constant-width font with surrounding quotes is used for filenames and
pathnames.  Example: ``The distribution is installed under the
@file{/usr/local/} directory.''

@item @samp{c}
Constant-width font with surrounding quotes is also used to indicate
character sequences.  Example: ``To specify a wildcard, use the @samp{%}
character.''

@item @emph{italic}
Italic font is used for emphasis, @emph{like this}.

@item @strong{boldface}
Boldface font is used for access privilege names (for example, ``do not grant
the @strong{process} privilege lightly'') and occasionally to convey
@strong{especially strong emphasis}.
@end table

When commands are shown that are meant to be executed by a particular
program, the program is indicated by a prompt shown before the command.  For
example, @code{shell>} indicates a command that you execute from your login
shell, and @code{mysql>} indicates a command that you execute from the
@code{mysql} client program:

@example
shell> type a shell command here
mysql> type a mysql command here
@end example

Shell commands are shown using Bourne shell syntax.  If you are using a
@code{csh}-style shell, you may need to issue commands slightly differently.
For example, the sequence to set an environment variable and run a command
looks like this in Bourne shell syntax:

@example
shell> VARNAME=value some_command
@end example

@cindex shell syntax
@cindex command syntax

For @code{csh}, you would execute the sequence like this:

@example
shell> setenv VARNAME value
shell> some_command
@end example

Often database, table, and column names must be substituted into commands.  To
indicate that such substitution is necessary, this manual uses
@code{db_name}, @code{tbl_name} and @code{col_name}.  For example, you might
see a statement like this:

@example
mysql> SELECT col_name FROM db_name.tbl_name;
@end example

This means that if you were to enter a similar statement, you would supply
your own database, table, and column names, perhaps like this:

@example
mysql> SELECT author_name FROM biblio_db.author_list;
@end example

SQL keywords are not case-sensitive and may be written in uppercase or
lowercase.  This manual uses uppercase.

In syntax descriptions, square brackets (@samp{[} and @samp{]}) are used
to indicate optional words or clauses.  For example, in the following
statement, @code{IF EXISTS} is optional:

@example
DROP TABLE [IF EXISTS] tbl_name
@end example

When a syntax element consists of a number of alternatives, the alternatives
are separated by vertical bars (@samp{|}).  When one member from a set of
choices @strong{may} be chosen, the alternatives are listed within square
brackets (@samp{[} and @samp{]}):

@example
TRIM([[BOTH | LEADING | TRAILING] [remstr] FROM] str)
@end example

When one member from a set of choices @strong{must} be chosen, the
alternatives are listed within braces (@samp{@{} and @samp{@}}):

@example
@{DESCRIBE | DESC@} tbl_name @{col_name | wild@}
@end example


@node What-is, What is MySQL AB, Manual-info, Introduction
@section What Is MySQL?

@cindex MySQL, defined
@cindex MySQL, introduction

@code{MySQL}, the most popular @code{Open Source} SQL database, is
developed and provided by @code{MySQL AB}.  @code{MySQL AB} is a
commercial company that builds its business providing services around
the @code{MySQL} database.
@xref{What is MySQL AB}.

The @code{MySQL} web site (@uref{http://www.mysql.com/})
provides the latest information about @code{MySQL} software and
@code{MySQL AB}.

@table @asis

@item @code{MySQL} is a database management system.

A database is a structured collection of data.  It may be anything from a
simple shopping list to a picture gallery or the vast amounts of
information in a corporate network.  To add, access, and process data
stored in a computer database, you need a database management system
such as @code{MySQL} Server.  Since computers are very good at handling large
amounts of data, database management plays a central role in computing,
as stand-alone utilities, or as parts of other applications.

@cindex databases, defined

@item MySQL is a relational database management system.

A relational database stores data in separate tables rather than putting
all the data in one big storeroom.  This adds speed and flexibility.
The tables are linked by defined relations making it possible to combine
data from several tables on request. The @code{SQL} part of
``@code{MySQL}'' stands for ``@code{Structured Query Language}''@-the
most common standardised language used to access databases.

@cindex relational databases, defined
@cindex SQL, defined
@item MySQL software is @code{Open Source}.

@code{Open Source} means that it is possible for anyone to use and modify.
Anybody can download the @code{MySQL} software from the Internet and use it
without paying anything.  Anybody so inclined can study the source code
and change it to fit their needs.  The @code{MySQL} software uses the
@code{GPL} (@code{GNU General Public License}),
@uref{http://www.gnu.org/licenses/}, to define what you
may and may not do with the software in different situations.
If you feel uncomfortable with the @code{GPL} or need to embed
@code{MySQL} code into a commercial application you can buy a
commercially licensed version from us.
@xref{MySQL licenses}.

@cindex @code{Open Source}, defined
@cindex General Public License

@item Why use the MySQL Database Server?

The @code{MySQL Database Server} is very fast, reliable, and easy to use.
If that is what you are looking for, you should give it a try.
@code{MySQL Server} also has a practical set of features developed in
close cooperation with our users.  You can find a performance comparison
of @code{MySQL Server} to some other database managers on our benchmark page.
@xref{MySQL Benchmarks}.

@code{MySQL Server} was originally developed to handle large databases
much faster than existing solutions and has been successfully used in
highly demanding production environments for several years.  Though
under constant development, @code{MySQL Server} today offers a rich and
useful set of functions.  Its connectivity, speed, and security make
@code{MySQL Server} highly suited for accessing databases on the Internet.

@item The technical features of MySQL Server

For advanced technical information, see @ref{Reference}.
The @code{MySQL Database Software} is a client/server system that consists
of a multi-threaded @code{SQL} server that supports different backends,
several different client programs and libraries, administrative tools,
and a wide range of programming interfaces (@code{API}s).

We also provide @code{MySQL Server} as a multi-threaded library which you
can link into your application to get a smaller, faster, easier-to-manage
product.

@item There is a large amount of contributed MySQL software available.

It is very likely that you will find that your favorite application or
language already supports the @code{MySQL Database Server}.

@end table

@cindex pronunciation, MySQL
@cindex MySQL, pronunciation
@cindex goals of MySQL
The official way to pronounce @code{MySQL} is ``My Ess Que Ell'' (not
``my sequel''), but we don't mind if you pronounce it as ``my sequel''
or in some other localised way.


@menu
* History::                     History of MySQL
* Features::                    The Main Features of MySQL
* Stability::                   How Stable Is MySQL?
* Table size::                  How Big Can MySQL Tables Be?
* Year 2000 compliance::        Year 2000 Compliance
@end menu

@node History, Features, What-is, What-is
@subsection History of MySQL

@cindex MySQL history
@cindex history of MySQL
@cindex MySQL name
@cindex My, derivation

We once started out with the intention of using @code{mSQL} to connect to our
tables using our own fast low-level (ISAM) routines. However, after some
testing we came to the conclusion that @code{mSQL} was not fast enough nor
flexible enough for our needs.  This resulted in a new SQL interface to our
database but with almost the same API interface as @code{mSQL}.  This API was
chosen to ease porting of third-party code.

The derivation of the name @code{MySQL} is not perfectly clear.  Our base
directory and a large number of our libraries and tools have had the prefix
``my'' for well over 10 years.  However, Monty's daughter (some years younger)
is also named My.  Which of the two gave its name to @code{MySQL} is
still a mystery, even for us.


@node Features, Stability, History, What-is
@subsection The Main Features of MySQL

@cindex main features of MySQL
@cindex features of MySQL

The following list describes some of the important characteristics
of the @code{MySQL Database Software}. @xref{MySQL 4.0 In A Nutshell}.

@c This list is too technical and should be divided into one feature
@c list comparable to commercial competition and a very technical on
@c with max limits (from crash-me) and so on.
@table @asis
@item Internals and Portability
@itemize @bullet
@item
Written in C and C++.  Tested with a broad range of different compilers.
@item
No memory leaks.  The @code{MySQL} code has been tested with Purify,
a commercial memory leakage detector.
@item
Works on many different platforms.  @xref{Which OS}.
@item
Uses GNU Automake (1.4), Autoconf (Version 2.52 or newer), and Libtool for
portability.
@item
APIs for C, C++, Eiffel, Java, Perl, PHP, Python, and Tcl.  @xref{Clients}.
@item
Fully multi-threaded using kernel threads.  This means it can easily
use multiple CPUs if available.
@item
Very fast B-tree disk tables with index compression.
@item
A very fast thread-based memory allocation system.
@item
Very fast joins using an optimised one-sweep multi-join.
@item
In-memory hash tables which are used as temporary tables.
@item
SQL functions are implemented through a highly optimised class library and
should be as fast as possible!  Usually there isn't any memory allocation
at all after query initialisation.
@end itemize

@item Column Types
@itemize @bullet
@item
Many column types: signed/unsigned integers 1, 2, 3, 4, and 8 bytes
long, @code{FLOAT}, @code{DOUBLE}, @code{CHAR}, @code{VARCHAR},
@code{TEXT}, @code{BLOB}, @code{DATE}, @code{TIME}, @code{DATETIME},
@code{TIMESTAMP}, @code{YEAR}, @code{SET}, and @code{ENUM} types.
@xref{Column types}.
@item
Fixed-length and variable-length records.
@item
@cindex default values
All columns have default values.
You can use @code{INSERT} to insert a subset of a table's columns; those
columns that are not explicitly given values are set to their default
values.
@end itemize

@item Commands and Functions
@itemize @bullet
@item
Full operator and function support in the @code{SELECT} and @code{WHERE}
parts of queries.  For example:

@example
mysql> SELECT CONCAT(first_name, " ", last_name)
    -> FROM tbl_name
    -> WHERE income/dependents > 10000 AND age > 30;
@end example

@item
Full support for SQL @code{GROUP BY} and
@code{ORDER BY} clauses.  Support
for group functions (@code{COUNT()},
@code{COUNT(DISTINCT ...)},
@code{AVG()}, @code{STD()},
@code{SUM()}, @code{MAX()}, and @code{MIN()}).
@item
Support for @code{LEFT OUTER JOIN} and @code{RIGHT OUTER JOIN} with ANSI
SQL and ODBC syntax.
@item
Aliases on tables and columns are allowed as in the SQL92 standard.
@item
@code{DELETE}, @code{INSERT}, @code{REPLACE}, and @code{UPDATE} return
the number of rows that were changed (affected).  It is possible to return
the number of rows matched instead by setting a flag when connecting to the
server.
@item
The @code{MySQL}-specific @code{SHOW} command can be used to retrieve
information about databases, tables, and indexes.  The @code{EXPLAIN} command
can be used to determine how the optimiser resolves a query.
@item
Function names do not clash with table or column names.  For example,
@code{ABS} is a valid column name.  The only restriction is that for a
function call, no spaces are allowed between the function name and the
@samp{(} that follows it.  @xref{Reserved words}.
@item
You can mix tables from different databases in the same query (as of
Version 3.22).
@end itemize

@item Security
@itemize @bullet
@item
A privilege and password system that is very flexible and secure, and
allows host-based verification.  Passwords are secure because all
password traffic is encrypted when you connect to a server.
@end itemize

@item Scalability and Limits
@itemize @bullet
@item
Handles large databases.  We are using @code{MySQL Server} with some
databases that contain 50 million records and we know of users that
use @code{MySQL Server} with 60,000 tables and about 5,000,000,000 rows.
@item
Up to 32 indexes per table are allowed.  Each index may consist of 1 to 16
columns or parts of columns.  The maximum index width is 500 bytes
(this may be changed when compiling @code{MySQL Server}).
An index may use a prefix of a @code{CHAR} or @code{VARCHAR} field.
@end itemize

@item Connectivity
@itemize @bullet
@item
Clients may connect to the @code{MySQL} server using TCP/IP Sockets,
Unix Sockets (Unix), or Named Pipes (NT).
@item
@code{ODBC} (Open-DataBase-Connectivity) support for Win32 (with source).
All ODBC 2.5 functions and many others.  For example, you can use
MS Access to connect to your @code{MySQL} server.  @xref{ODBC}.
@end itemize

@item Localisation
@itemize @bullet
@item
The server can provide error messages to clients in many languages.
@xref{Languages}.
@item
Full support for several different character sets, including
ISO-8859-1 (Latin1), german, big5, ujis, and more.  For example,
the Scandinavian characters 'å', 'ä' and 'ö' are allowed in table
and column names.
@item
All data is saved in the chosen character set.  All comparisons for normal
string columns are case-insensitive.
@item
Sorting is done according to the chosen character set (the Swedish
way by default).  It is possible to change this when the @code{MySQL}
server is started.  To see an example of very advanced sorting, look
at the Czech sorting code.  @code{MySQL Server} supports many different
character sets that can be specified at compile and runtime.
@end itemize

@item Clients and Tools
@itemize @bullet
@item
Includes @code{myisamchk}, a very fast utility for table checking,
optimisation, and repair.  All of the functionality of @code{myisamchk}
is also available through the SQL interface as well.
@xref{MySQL Database Administration}.
@item
All @code{MySQL} programs can be invoked with the @code{--help} or @code{-?}
options to obtain online assistance.
@end itemize

@end table


@node Stability, Table size, Features, What-is
@subsection How Stable Is MySQL?

@cindex stability

This section addresses the questions ``@emph{How stable is MySQL Server?}''
and ``@emph{Can I depend on MySQL Server in this project?}''
We will try to clarify these issues and answer some important
questions that concern many potential users. The information in this
section is based on data gathered from the mailing list, which is
very active in identifying problems as well as reporting types of use.

Original code stems back from the early '80s, providing a stable code
base, and the ISAM table format remains backward-compatible.
At TcX, the predecessor of @code{MySQL AB}, @code{MySQL} code has worked
in projects since mid-1996, without any problems.
When the @code{MySQL Database Software} was released to a wider public,
we noticed that there were
some pieces of ``untested code'' that were quickly found by the new
users who made different types of queries from us. Each new release
has had fewer portability problems (even though each new release has
had many new features).

Each release of the @code{MySQL Server} has been usable. There have only
been problems
when users try code from the ``gray zones.''  Naturally, new users
don't know what the gray zones are; this section attempts to indicate
those that are currently known.
The descriptions mostly deal with Version 3.23 of @code{MySQL Server}.
All known and reported bugs are fixed in the latest version, with the
exception of those listed in the bugs section, which are things that
are design-related.  @xref{Bugs}.

The @code{MySQL Server} design is multi-layered with independent modules.
Some of the newer modules are listed here with an indication of how
well-tested each of them is:

@cindex modules, list of

@table @strong
@item Replication -- Gamma
Large server clusters using replication are in production use, with
good results. Work on enhanced replication features is continuing
in @code{MySQL} 4.0.

@item @code{InnoDB} tables -- Stable (in 3.23 from 3.23.49)
The @code{InnoDB} transactional table handler has now been declared
stable in the @code{MySQL} 3.23 tree, starting from version 3.23.49.
@code{InnoDB} is being used in large, heavy-load production systems.

@item @code{BDB} tables -- Gamma
The @code{Berkeley DB} code is very stable, but we are still improving
the @code{BDB} transactional table handler interface in
@code{MySQL Server}, so it will take some time before this is as well
tested as the other table types.

@item @code{FULLTEXT} -- Beta
Full-text search works but is not yet widely used.
Important enhancements are being implemented for @code{MySQL} 4.0.

@item @code{MyODBC 2.50} (uses ODBC SDK 2.5) -- Gamma
Increasingly in wide use. Some issues brought up appear to be
application-related and independent of the ODBC driver or underlying
database server.

@item Automatic recovery of @code{MyISAM} tables -- Gamma
This status only regards the new code in the @code{MyISAM} table
handler that checks if the table was closed properly on open and
executes an automatic check/repair of the table if it wasn't.

@item Bulk-insert -- Alpha
New feature in @code{MyISAM} tables in @code{MySQL} 4.0 for faster
insert of many rows.

@item Locking -- Gamma
This is very system-dependent.  On some systems there are big problems
using standard OS locking (@code{fcntl()}).  In these cases, you should
run @code{mysqld} with the @code{--skip-locking} flag.
Problems are known to occur on some Linux systems, and on SunOS when
using NFS-mounted filesystems.

@end table

@code{MySQL AB} provides high-quality support for paying customers,
but the @code{MySQL} mailing list usually provides answers to common
questions. Bugs are usually fixed right away with a patch; for serious
bugs, there is almost always a new release.


@node Table size, Year 2000 compliance, Stability, What-is
@subsection How Big Can MySQL Tables Be?

@cindex tables, maximum size
@cindex size of tables
@cindex operating systems, file-size limits
@cindex limits, file-size
@cindex files, size limits

@code{MySQL} Version 3.22 has a 4G limit on table size. With the new
@code{MyISAM} table type in @code{MySQL} Version 3.23, the maximum table
size is pushed up to 8 million terabytes (2 ^ 63 bytes).

Note, however, that operating systems have their own file-size
limits. Here are some examples:

@multitable @columnfractions .30 .50
@item @strong{Operating System} @tab @strong{File-Size Limit}
@item Linux-Intel 32 bit        @tab 2G, 4G or more, depends on Linux version
@item Linux-Alpha               @tab 8T (?)
@item Solaris 2.5.1             @tab 2G (possible 4G with patch)
@item Solaris 2.6               @tab 4G (can be changed with flag)
@item Solaris 2.7 Intel         @tab 4G
@item Solaris 2.7 UltraSPARC    @tab 512G
@end multitable

On Linux 2.2 you can get bigger tables than 2G by using the LFS patch for
the ext2 filesystem.  On Linux 2.4 patches also exist for ReiserFS
to get support for big files.

This means that the table size for @code{MySQL} databases is normally
limited by the operating system.

By default, @code{MySQL} tables have a maximum size of about 4G.  You can
check the maximum table size for a table with the @code{SHOW TABLE STATUS}
command or with the @code{myisamchk -dv table_name}.
@xref{SHOW}.

If you need bigger tables than 4G (and your operating system supports
this), you should set the @code{AVG_ROW_LENGTH} and @code{MAX_ROWS}
parameter when you create your table.  @xref{CREATE TABLE}.  You can
also set these later with @code{ALTER TABLE}. @xref{ALTER TABLE}.

If your big table is going to be read-only, you could use
@code{myisampack} to merge and compress many tables to one.
@code{myisampack} usually compresses a table by at least 50%, so you can
have, in effect, much bigger tables.  @xref{myisampack, ,
@code{myisampack}}.

You can go around the operating system file limit for @code{MyISAM} data
files by using the @code{RAID} option. @xref{CREATE TABLE}.

Another solution can be the included @code{MERGE} library, which allows
you to handle a collection of identical tables as one.
@xref{MERGE, , @code{MERGE} tables}.


@node Year 2000 compliance,  , Table size, What-is
@subsection Year 2000 Compliance

@cindex Year 2000 compliance
@cindex compliance, Y2K
@cindex date functions, Y2K compliance

The @code{MySQL Server} itself has no problems with Year 2000 (Y2K)
compliance:

@itemize @bullet
@item
@code{MySQL Server} uses Unix time functions and has no problems with dates
until @code{2069}; all 2-digit years are regarded to be in the range
@code{1970} to @code{2069}, which means that if you store @code{01} in a
@code{year} column, @code{MySQL Server} treats it as @code{2001}.

@item
All @code{MySQL} date functions are stored in one file, @file{sql/time.cc},
and are coded very carefully to be year 2000-safe.

@item
In @code{MySQL} Version 3.22 and later, the new @code{YEAR} column type
can store years @code{0} and @code{1901} to @code{2155} in 1 byte and
display them using 2 or 4 digits.
@end itemize

You may run into problems with applications that use @code{MySQL Server}
in a way that is not Y2K-safe.  For example, many old applications store
or manipulate years using 2-digit values (which are ambiguous) rather than
4-digit values.  This problem may be compounded by applications that use
values such as @code{00} or @code{99} as ``missing'' value indicators.

Unfortunately, these problems may be difficult to fix because different
applications may be written by different programmers, each of whom may
use a different set of conventions and date-handling functions.

Here is a simple demonstration illustrating that @code{MySQL Server}
doesn't have any problems with dates until the year 2030:

@example
mysql> DROP TABLE IF EXISTS y2k;
Query OK, 0 rows affected (0.01 sec)

mysql> CREATE TABLE y2k (date DATE,
    ->                   date_time DATETIME,
    ->                   time_stamp TIMESTAMP);
Query OK, 0 rows affected (0.00 sec)

mysql> INSERT INTO y2k VALUES
    -> ("1998-12-31","1998-12-31 23:59:59",19981231235959),
    -> ("1999-01-01","1999-01-01 00:00:00",19990101000000),
    -> ("1999-09-09","1999-09-09 23:59:59",19990909235959),
    -> ("2000-01-01","2000-01-01 00:00:00",20000101000000),
    -> ("2000-02-28","2000-02-28 00:00:00",20000228000000),
    -> ("2000-02-29","2000-02-29 00:00:00",20000229000000),
    -> ("2000-03-01","2000-03-01 00:00:00",20000301000000),
    -> ("2000-12-31","2000-12-31 23:59:59",20001231235959),
    -> ("2001-01-01","2001-01-01 00:00:00",20010101000000),
    -> ("2004-12-31","2004-12-31 23:59:59",20041231235959),
    -> ("2005-01-01","2005-01-01 00:00:00",20050101000000),
    -> ("2030-01-01","2030-01-01 00:00:00",20300101000000),
    -> ("2050-01-01","2050-01-01 00:00:00",20500101000000);
Query OK, 13 rows affected (0.01 sec)
Records: 13  Duplicates: 0  Warnings: 0

mysql> SELECT * FROM y2k;
+------------+---------------------+----------------+
| date       | date_time           | time_stamp     |
+------------+---------------------+----------------+
| 1998-12-31 | 1998-12-31 23:59:59 | 19981231235959 |
| 1999-01-01 | 1999-01-01 00:00:00 | 19990101000000 |
| 1999-09-09 | 1999-09-09 23:59:59 | 19990909235959 |
| 2000-01-01 | 2000-01-01 00:00:00 | 20000101000000 |
| 2000-02-28 | 2000-02-28 00:00:00 | 20000228000000 |
| 2000-02-29 | 2000-02-29 00:00:00 | 20000229000000 |
| 2000-03-01 | 2000-03-01 00:00:00 | 20000301000000 |
| 2000-12-31 | 2000-12-31 23:59:59 | 20001231235959 |
| 2001-01-01 | 2001-01-01 00:00:00 | 20010101000000 |
| 2004-12-31 | 2004-12-31 23:59:59 | 20041231235959 |
| 2005-01-01 | 2005-01-01 00:00:00 | 20050101000000 |
| 2030-01-01 | 2030-01-01 00:00:00 | 20300101000000 |
| 2050-01-01 | 2050-01-01 00:00:00 | 00000000000000 |
+------------+---------------------+----------------+
13 rows in set (0.00 sec)

@end example

This shows that the @code{DATE} and @code{DATETIME} types will not
give any problems with future dates (they handle dates until the year
9999).

The @code{TIMESTAMP} type, which is used to store the current time, has a
range up to only @code{2030-01-01}.  @code{TIMESTAMP} has a range of
@code{1970} to @code{2030} on 32-bit machines (signed value).  On 64-bit
machines it handles times up to @code{2106} (unsigned value).

Even though @code{MySQL Server} is Y2K-compliant, it is your responsibility
to provide unambiguous input.  See @ref{Y2K issues} for @code{MySQL Server}'s
rules for dealing with ambiguous date input data (data containing 2-digit
year values).



@node What is MySQL AB, Licensing and Support, What-is, Introduction
@section What Is MySQL AB?

@cindex MySQL AB, defined

@code{MySQL AB} is the company of the @code{MySQL} founders and main
developers. @code{MySQL AB} was originally established in Sweden by
David Axmark, Allan Larsson, and Michael @code{Monty} Widenius.

All the developers of the @code{MySQL} server are employed by the company.
We are a virtual organisation with people in a dozen countries around
the world. We communicate extensively over the Net every day with each
other and with our users, supporters and partners.

We are dedicated to developing the @code{MySQL} software and spreading
our database to new users. @code{MySQL AB} owns the copyright to the
@code{MySQL} source code, the @code{MySQL} logo and trademark, and this
manual. @xref{What-is}.

@menu
* MySQL AB business model and services::  The Business Model and Services of MySQL AB
* Contact information::         Contact Information
@end menu

The @code{MySQL} core values show our dedication to @code{MySQL} and
@code{Open Source}.

We want the @code{MySQL Database Software} to be:
@itemize @bullet
@item
The best and the most widely used database in the world.
@item
Available and affordable for all.
@item
Easy to use.
@item
Continuously improving while remaining fast and safe.
@item
Fun to use and improve.
@item
Free from bugs.
@end itemize

@code{MySQL AB} and the people at @code{MySQL AB}:
@itemize @bullet
@item
Promote @code{Open Source} philosophy and support the
@code{Open Source} community.
@item
Aim to be good citizens.
@item
Prefer partners that share our values and mind-set.
@item
Answer e-mail and provide support.
@item
Are a virtual company, networking with others.
@item
Work against software patents.
@end itemize

The @code{MySQL} web site (@uref{http://www.mysql.com/})
provides the latest information about @code{MySQL} and @code{MySQL AB}.


@node MySQL AB business model and services, Contact information, What is MySQL AB, What is MySQL AB
@subsection The Business Model and Services of MySQL AB

One of the most common questions we encounter is:
``@emph{How can you make a living from something you give away for free?}''
This is how.

@code{MySQL AB} makes money on support, services, commercial licenses,
and royalties, and we use these revenues to fund product development
and to expand the @code{MySQL} business.

@menu
* Business Services Support::   Support
* Business Services Training::  Training and Certification
* Business Services Consulting::  Consulting
* Business Services Commercial licenses::  Commercial Licenses
* Business Services Partnering::  Partnering
* Business Services Advertising::  Advertising
@end menu

The company has been profitable since its inception. In October 2001,
we accepted venture financing from leading Scandinavian investors and
a handful of business angels. This investment is used to solidify our
business model and build a basis for sustainable growth.


@node Business Services Support, Business Services Training, MySQL AB business model and services, MySQL AB business model and services
@subsubsection Support
@code{MySQL AB} is run and owned by the founders and main developers of
the @code{MySQL} database. The developers are committed to giving support
to customers and other users in order to stay in touch with their needs
and problems. All our support is given by qualified developers. Really
tricky questions are answered by Michael @code{Monty} Widenius, principal
author of the @code{MySQL Server}.
@xref{Support}.

To order support at various levels, please visit the order section at
@uref{https://order.mysql.com/} or contact our sales staff at
@email{sales@@mysql.com}.


@node Business Services Training, Business Services Consulting, Business Services Support, MySQL AB business model and services
@subsubsection Training and Certification
@cindex training
@cindex MySQL training
@cindex certification
@cindex MySQL certification
@code{MySQL AB} delivers @code{MySQL} and related training worldwide.
We offer both open courses and in-house courses tailored to the
specific needs of your company. @code{MySQL Training} is also available
through our partners, the @code{Authorised MySQL Training Centers}.

Our training material uses the same example databases as our
documentation and our sample applications, and it is always updated
to reflect the latest @code{MySQL} version. Our trainers are backed by
the development team to guarantee the quality of the training and the
continuous development of the course material. This also ensures
that no questions raised during the courses remain unanswered.

Attending our training courses will enable you to achieve your goals
related to your @code{MySQL} applications. You will also:
@itemize @bullet
@item
Save time.
@item
Improve the performance of your application(s).
@item
Reduce or eliminate the need for additional hardware, decreasing cost.
@item
Enhance security.
@item
Increase customers' and co-workers' satisfaction.
@item
Prepare yourself for @code{MySQL Certification}.
@end itemize

If you are interested in our training as a potential participant or
as a training partner, please visit the training section at
@uref{http://www.mysql.com/training/} or contact us at:
@email{training@@mysql.com}.

We plan to release the @code{MySQL Certification Program} in 2002. For
details see @uref{http://www.mysql.com/training/certification.html}.
If you would like to be kept informed about the
@code{MySQL Certification Program},
please e-mail @email{certification@@mysql.com}.


@node Business Services Consulting, Business Services Commercial licenses, Business Services Training, MySQL AB business model and services
@subsubsection Consulting
@cindex consulting
@cindex MySQL consulting
@code{MySQL AB} and its @code{Authorised Partners} offer consulting
services to users of @code{MySQL Server} and to those who embed
@code{MySQL Server} in their own software, all over the world.

Our consultants can help you design and tune your databases, construct
efficient queries, tune your platform for optimal performance, resolve
migration issues, set up replication, build robust transactional
applications, and more.
We also help customers embed @code{MySQL Server} in their products and
applications for large-scale deployment.

Our consultants work in close collaboration with our development team,
which ensures the technical quality of our professional services.
Consulting assignments range from 2-day power-start sessions to
projects that span weeks and months. Our expertise not only covers
@code{MySQL Server}, but also extends into programming and scripting
languages such as PHP, Perl, and more.

If you are interested in our consulting services or want to become a
consulting partner, please visit the consulting section of our web site
at @uref{http://www.mysql.com/consulting/} or contact our consulting
staff at @email{consulting@@mysql.com}.


@node Business Services Commercial licenses, Business Services Partnering, Business Services Consulting, MySQL AB business model and services
@subsubsection Commercial Licenses
The @code{MySQL} database is released under the
@code{GNU General Public License} (@code{GPL}).
This means that the @code{MySQL} software can be used free of charge
under the @code{GPL}. If you do not want to be bound by the @code{GPL}
terms (like the requirement that your own application becomes @code{GPL}
as well), you may purchase a commercial license for the same product
from @code{MySQL AB} at @uref{https://order.mysql.com/}.
Since @code{MySQL AB} owns the copyright to the @code{MySQL} source code,
we are able to employ @code{Dual Licensing} which means that the same
product is available under @code{GPL} and under a commercial
license. This does not in any way affect the @code{Open Source}
commitment of @code{MySQL AB}. For details about when a commercial
license is required, please see @ref{MySQL licenses}.

We also sell commercial licenses of third-party @code{Open Source GPL}
software that adds value to @code{MySQL Server}. A good example is the
@code{InnoDB} transactional table handler that offers @code{ACID}
support, row-level locking, crash recovery, multi-versioning, foreign
key support, and more. @xref{InnoDB}.


@node Business Services Partnering, Business Services Advertising, Business Services Commercial licenses, MySQL AB business model and services
@subsubsection Partnering
@cindex partnering with MySQL AB
@code{MySQL AB} has a worldwide partner programme that covers training
courses, consulting & support, publications plus reselling and
distributing @code{MySQL} and related products. @code{MySQL AB Partners}
get visibility on the @uref{http://www.mysql.com/} web site and the right
to use special versions of the @code{MySQL} trademarks to identify their
products and promote their business.

If you are interested in becoming a @code{MySQL AB Partner}, please e-mail
@email{partner@@mysql.com}.

The word @code{MySQL} and the @code{MySQL} dolphin logo are trademarks of
@code{MySQL AB}. @xref{MySQL AB Logos and Trademarks}.
These trademarks represent a significant value that the @code{MySQL}
founders have built over the years.


@node Business Services Advertising,  , Business Services Partnering, MySQL AB business model and services
@subsubsection Advertising
The @code{MySQL} web site (@uref{http://www.mysql.com/}) is popular among
developers and users. In October 2001, we served 10 million page views.
Our visitors represent a group that makes purchase decisions and
recommendations for both software and hardware. Twelve percent of our
visitors authorise purchase decisions, and only nine percent are not
involved in purchase decisions at all. More than 65% have made one or
more online business purchase within the last half-year, and 70% plan
to make one in the next months.

If you are interested in placing banner ads on our web site,
@uref{http://www.mysql.com/}, please send an e-mail message to
@email{advertising@@mysql.com}.


@node Contact information,  , MySQL AB business model and services, What is MySQL AB
@subsection Contact Information

@cindex contact information
@cindex employment, contact information
@cindex licensing, contact information
@cindex advertising, contact information
@cindex employment with MySQL
@cindex jobs at MySQL

The @code{MySQL} web site (@uref{http://www.mysql.com/})
provides the latest information about @code{MySQL} and @code{MySQL AB}.

For press service and inquiries not covered in our News releases
(@uref{http://www.mysql.com/news/}), please send e-mail to
@email{press@@mysql.com}.

If you have a valid support contract with @code{MySQL AB}, you will
get timely, precise answers to your technical questions about the
@code{MySQL} software. For more information, see @ref{Support}.
You can order your support contract at
@uref{https://order.mysql.com/}, or send an e-mail message to
@email{sales@@mysql.com}.

For information about @code{MySQL} training, please visit the training
section at @uref{http://www.mysql.com/training/}. If you have
restricted access to the Internet, please contact the @code{MySQL AB}
training staff at @email{training@@mysql.com}.
@xref{Business Services Training}.

For information on the @code{MySQL Certification Program}, please see
@uref{http://www.mysql.com/training/certification.html}.
If you would like to be kept informed about the
@code{MySQL Certification Program}, please e-mail
@email{certification@@mysql.com}.
@xref{Business Services Training}.

If you're interested in consulting, please visit the consulting
section at @uref{http://www.mysql.com/consulting/}. If you have
restricted access to the Internet, please contact the @code{MySQL AB}
consulting staff at @email{consulting@@mysql.com}.
@xref{Business Services Consulting}.

Commercial licenses may be purchased online at
@uref{https://order.mysql.com/}. There you will also find information
on how to fax your purchase order to @code{MySQL AB}. If you have
questions regarding licensing or you want a quote for a high-volume
license deal, please fill in the contact form on our web site
(@uref{http://www.mysql.com/}) or send an e-mail message
to @email{licensing@@mysql.com} (for licensing questions) or to
@email{sales@@mysql.com} (for sales inquiries).
@xref{MySQL licenses}.

If you represent a business that is interested in partnering with
@code{MySQL AB}, please send e-mail to @email{partner@@mysql.com}.
@xref{Business Services Partnering}.

If you are interested in placing a banner advertisement on the
@code{MySQL} web site (@uref{http://www.mysql.com/}), please send
e-mail to @email{advertising@@mysql.com}.
@xref{Business Services Advertising}.

For more information on the @code{MySQL} trademark policy, refer to
@uref{http://www.mysql.com/company/trademark.html} or send e-mail to
@email{trademark@@mysql.com}.
@xref{MySQL AB Logos and Trademarks}.

If you are interested in any of the @code{MySQL AB} jobs listed in our
jobs section (@uref{http://www.mysql.com/development/jobs/}),
please send an e-mail message to @email{jobs@@mysql.com}.
Please do not send your CV as an attachment, but rather as plain text
at the end of your e-mail message.

For general discussion among our many users, please direct your
attention to the appropriate mailing list.
@xref{Questions}.

Reports of errors (often called bugs), as well as questions and
comments, should be sent to the mailing list at
@email{mysql@@lists.mysql.com}. If you have found a sensitive
security bug in the @code{MySQL Server}, please send an e-mail
to @email{security@@mysql.com}.
@xref{Bug reports}.

If you have benchmark results that we can publish, please
contact us at @email{benchmarks@@mysql.com}.

If you have any suggestions concerning additions or corrections to this
manual, please send them to the manual team at @email{docs@@mysql.com}.

For questions or comments about the workings or content of the
@code{MySQL} web site (@uref{http://www.mysql.com/}),
please send e-mail to @email{webmaster@@mysql.com}.

Questions about the @code{MySQL Portals}
(@uref{http://www.mysql.com/portal/})
may be sent to @email{portals@@mysql.com}.

@code{MySQL AB} has a privacy policy, which can be read at
@uref{http://www.mysql.com/company/privacy.html}.
For any queries regarding this policy, please e-mail
@email{privacy@@mysql.com}.

For all other inquires, please send e-mail to @email{info@@mysql.com}.


@node Licensing and Support, MySQL 4.0 In A Nutshell, What is MySQL AB, Introduction
@section MySQL Support and Licensing

@cindex licensing terms
@cindex support terms

This section describes @code{MySQL} support and licensing arrangements.

@menu
* Support::                     Support Offered by MySQL AB
* Copyright::                   Copyrights and Licenses Used by MySQL
* MySQL licenses::              MySQL Licenses
* MySQL AB Logos and Trademarks::  MySQL AB Logos and Trademarks
@end menu


@node Support, Copyright, Licensing and Support, Licensing and Support
@subsection Support Offered by MySQL AB

@cindex support, types
@cindex types of support
@cindex commercial support, types
@cindex costs, support
@cindex licensing costs
@cindex support costs
@cindex prices, support

Technical support from @code{MySQL AB} means individualised answers
to your unique problems direct from the software engineers who code
the @code{MySQL} database engine.

We try to take a broad and inclusive view of technical support. Almost
any problem involving @code{MySQL} software is important to us if it's
important to you.
Typically customers seek help on how to get different commands and
utilities to work, remove performance bottlenecks, restore crashed
systems, understand operating system or networking impacts on @code{MySQL},
set up best practices for backup and recovery, utilise @code{API}s, etc.
Our support covers only the @code{MySQL} server and our own utilities,
not third-party products that access the @code{MySQL} server, though we
try to help with these where we can.

Detailed information about our various support options is given at
@uref{https://order.mysql.com/}, where support contracts can also be
ordered online. If you have restricted access to the Internet, contact
our sales staff at @email{sales@@mysql.com}.

Technical support is like life insurance. You can live happily
without it for years, but when your hour arrives it becomes
critically important, yet it's too late to buy it!
If you use @code{MySQL} Server for important applications and encounter
sudden troubles,  it might take too long to figure out all the answers
yourself. You may need immediate access to the most experienced
@code{MySQL} troubleshooters available, those employed by @code{MySQL AB}.


@node Copyright, MySQL licenses, Support, Licensing and Support
@subsection Copyrights and Licenses Used by MySQL

@cindex copyrights
@cindex licenses

@code{MySQL AB} owns the copyright to the @code{MySQL} source code,
the @code{MySQL} logos and trademarks and this manual.
@xref{What is MySQL AB}.
Several different licenses are relevant to the @code{MySQL}
distribution:

@enumerate
@item
All the @code{MySQL}-specific source in the server, the @code{mysqlclient}
library and the client, as well as the @code{GNU} @code{readline} library
is covered by the @code{GNU General Public License}.
@xref{GPL license}.
The text of this license can also be found as the file @file{COPYING}
in the distributions.

@item
The @code{GNU} @code{getopt} library is covered by the
@code{GNU Lesser General Public License}.
@xref{LGPL license}.

@item
Some parts of the source (the @code{regexp} library) are covered
by a Berkeley-style copyright.

@item
Older versions of @code{MySQL} (3.22 and earlier) are subject to a
more strict license
(@uref{http://www.mysql.com/support/arrangements/mypl.html}).
See the documentation of the specific version for information.

@item
The manual is currently @strong{not} distributed under a @code{GPL}-style license.
Use of the manual is subject to the following terms:
@itemize @bullet
@item
Conversion to other formats is allowed, but the actual content
may not be altered or edited in any way.
@item
You may create a printed copy for your own personal use.
@item
For all other uses, such as selling printed copies or using
(parts of) the manual in another publication, prior written
agreement from @code{MySQL AB} is required.
@end itemize
Please e-mail @email{docs@@mysql.com} for more information or
if you are interested in doing a translation.
@end enumerate

For information about how the @code{MySQL} licenses work in practice,
please refer to @ref{MySQL licenses}.
Also see @ref{MySQL AB Logos and Trademarks}.


@node MySQL licenses, MySQL AB Logos and Trademarks, Copyright, Licensing and Support
@subsection MySQL Licenses

@cindex licensing policy
@cindex technical support, licensing
@cindex support, licensing
@cindex General Public License, MySQL
@cindex GPL, MySQL
@cindex licensing, examples
@cindex selling products
@cindex products, selling

The @code{MySQL} software is released under the
@code{GNU General Public License} (@code{GPL}),
which probably is the best known @code{Open Source} license.
The formal terms of the @code{GPL} license can be found at
@uref{http://www.gnu.org/licenses/}.
See also @uref{http://www.gnu.org/licenses/gpl-faq.html} and
@uref{http://www.gnu.org/philosophy/enforcing-gpl.html}.

Since the @code{MySQL} software is released under the @code{GPL},
it may often be used for free, but for certain uses you may want
or need to buy commercial licenses from @code{MySQL AB} at
@uref{https://order.mysql.com/}.

Older versions of @code{MySQL} (3.22 and earlier) are subject to a
more strict license
(@uref{http://www.mysql.com/support/arrangements/mypl.html}).
See the documentation of the specific version for information.

Please note that the use of the @code{MySQL} software under commercial
license, @code{GPL}, or the old @code{MySQL} license does not
automatically give you the right to use @code{MySQL AB} trademarks.
@xref{MySQL AB Logos and Trademarks}.


@menu
* Using the MySQL software under a commercial license::  Using the MySQL Software Under a Commercial License
* Using the MySQL software for free under GPL::  Using the MySQL Software for Free Under GPL
@end menu

@node Using the MySQL software under a commercial license, Using the MySQL software for free under GPL, MySQL licenses, MySQL licenses
@subsubsection Using the MySQL Software Under a Commercial License

The @code{GPL} license is contagious in the sense that when a program
is linked to a @code{GPL} program the resulting product must also be
released under @code{GPL} lest you break the license terms and forfeit
your right to use the @code{GPL} program altogether.

You need a commercial license:

@itemize @bullet
@item
When you link a program with code from the @code{MySQL} software or from
@code{GPL} released clients and don't want the resulting product to be
@code{GPL}, maybe because you want to build a commercial product or keep
the added non-@code{GPL} code closed source for other reasons. When
purchasing commercial licenses, you are not using the @code{MySQL} software
under @code{GPL} even though it's the same code.

@item
When you distribute a non-@code{GPL} application that @strong{only} works with the
@code{MySQL} software and ship it with the @code{MySQL} software. This type
of solution is actually considered to be linking even if it's done over
a network.

@item
When you distribute copies of the @code{MySQL} software without providing
the source code as required under the @code{GPL} license.

@item
When you want to support the further development of the @code{MySQL}
database even if you don't formally need a commercial license.
Purchasing support directly from @code{MySQL AB} is another good way
of contributing to the development of the @code{MySQL} software, with
immediate advantages for you.
@xref{Support}.
@end itemize

If you require a license, you will need one for each installation of the
@code{MySQL} software. This covers any number of CPUs on a machine, and there
is no artificial limit on the number of clients that connect to the server
in any way.

To purchase commercial licenses and support, please visit the order section
of our web site at @uref{https://order.mysql.com/}. If you have special
licensing needs or you have restricted access to the Internet, please contact
our sales staff at @email{sales@@mysql.com}.


@node Using the MySQL software for free under GPL,  , Using the MySQL software under a commercial license, MySQL licenses
@subsubsection Using the MySQL Software for Free Under GPL

@cindex licensing, free
@cindex free licensing

You can use the @code{MySQL} software for free under the @code{GPL}:

@itemize @bullet
@item
When you link a program with code from the @code{MySQL} software and
release the resulting product under @code{GPL}.

@item
When you distribute the @code{MySQL} source code bundled with
other programs that are not linked to or dependent on @code{MySQL Server}
for their functionality even if you sell the distribution commercially.

@item
When using the @code{MySQL} software internally in your company.

@item
@cindex ISP services
@cindex services, ISP
@cindex services, web
@cindex Internet Service Providers
When you are an Internet Service Provider (ISPs) offering web
hosting with @code{MySQL} servers for your customers.
On the other hand, we do encourage people to use ISPs that have
@code{MySQL} support, as this will give them the confidence that if
they have some problem with the @code{MySQL} installation, their ISP
will in fact have the resources to solve the problem for them.

All ISPs that want to keep themselves up-to-date should subscribe
to our @code{announce} mailing list so that they can be aware of
critical issues that may be relevant for their @code{MySQL} installations.

Note that even if an ISP does not have a commercial license for
@code{MySQL Server}, they should at least give their customers read
access to the source of the @code{MySQL} installation so that the
customers can verify that it is patched correctly.

@item
@cindex web server, running
@cindex running a web server
When you use the @code{MySQL} Database Software  in conjunction with
a web server, you do not need a commercial license. This is true even
if you run a commercial web server that uses @code{MySQL Server}, because
you are not selling an embedded @code{MySQL} version yourself.  However,
in this case we would like you to purchase @code{MySQL} support because
the @code{MySQL} software is helping your enterprise.
@end itemize

If your use of @code{MySQL} database software does not require a commercial
license, we encourage you to purchase support from @code{MySQL AB} anyway.
This way you contribute toward @code{MySQL} development and also gain
immediate advantages for yourself. @xref{Support}.

If you use the @code{MySQL} database software in a commercial context
such that you profit by its use, we ask that you further the development
of the @code{MySQL} software by purchasing some level of support.  We feel
that if the @code{MySQL} database helps your business, it is reasonable to
ask that you help @code{MySQL AB}.
(Otherwise, if you ask us support questions, you are not only using
for free something into which we've put a lot a work, you're asking
us to provide free support, too.)


@node MySQL AB Logos and Trademarks,  , MySQL licenses, Licensing and Support
@subsection MySQL AB Logos and Trademarks

@cindex logos
@cindex trademarks

Many users of the @code{MySQL} database want to display the
@code{MySQL AB} dolphin logo on their web sites, books, or
boxed products. We welcome and encourage this, although it should be
noted that the word @code{MySQL} and the @code{MySQL} dolphin logo
are trademarks of @code{MySQL AB} and may only be used as stated in
our trademark policy at
@uref{http://www.mysql.com/company/trademark.html}.

@menu
* The Original MySQL logo::     The Original MySQL Logo
* MySQL Logos that may be Used Without Written Permission::  MySQL Logos that may be Used Without Written Permission
* When need MySQL logo permission::  When do you need a Written Permission to use MySQL Logos?
* MySQL AB Partnership Logos::  MySQL AB Partnership Logos
* Using MySQL word in presentations::  Using the word @code{MySQL} in Printed Text or Presentations
* Using MySQL word in company and product names::  Using the word @code{MySQL} in Company and Product Names
@end menu


@node The Original MySQL logo, MySQL Logos that may be Used Without Written Permission, MySQL AB Logos and Trademarks, MySQL AB Logos and Trademarks
@subsubsection The Original MySQL Logo

@c FIX: picture: MySQL logo

The @code{MySQL} dolphin logo was designed by the Finnish advertising
agency Priority in 2001.  The dolphin was chosen as a suitable symbol
for the @code{MySQL} database since it is a smart, fast, and lean animal,
effortlessly navigating oceans of data. We also happen to like dolphins.

The original @code{MySQL} logo may only be used by representatives of
@code{MySQL AB} and by those having a written agreement allowing them
to do so.


@node MySQL Logos that may be Used Without Written Permission, When need MySQL logo permission, The Original MySQL logo, MySQL AB Logos and Trademarks
@subsubsection MySQL Logos that may be Used Without Written Permission

@c FIX: pictures: powered by, works with, included - logos

We have designed a set of special @emph{Conditional Use} logos that may be
downloaded from our web site at
@uref{http://www.mysql.com/downloads/logos.html}
and used on third-party web sites without written permission from
@code{MySQL AB}.
The use of these logos is not entirely unrestricted but, as the name
implies, subject to our trademark policy that is also available on our
web site. You should read through the trademark policy if you plan to
use them. The requirements are basically:

@itemize @bullet
@item
Use the logo you need as displayed on the @uref{http://www.mysql.com/}
site. You may scale it to fit your needs, but not change colours or design,
or alter the graphics in any way.

@item
Make it evident that you, and not @code{MySQL AB}, are the creator and
owner of the site that displays the @code{MySQL} trademark.

@item
Don't use the trademark in a way that is detrimental to @code{MySQL AB}
or to the value of @code{MySQL AB} trademarks. We reserve the right to
revoke the right to use the @code{MySQL AB} trademark.

@item
If you use the trademark on a web site, make it clickable, leading directly
to @uref{http://www.mysql.com/}.

@item
If you are using the @code{MySQL} database under @code{GPL} in an
application, your application must be @code{Open Source} and
be able to connect to a @code{MySQL} server.
@end itemize

Contact us at @email{trademark@@mysql.com} to inquire about special
arrangements to fit your needs.

@node When need MySQL logo permission, MySQL AB Partnership Logos, MySQL Logos that may be Used Without Written Permission, MySQL AB Logos and Trademarks
@subsubsection When do you need a Written Permission to use MySQL Logos?

In the following cases you need a written permission from @code{MySQL AB}
before using @code{MySQL} logos:

@itemize @bullet
@item
When displaying any @code{MySQL AB} logo anywhere except on your web site.

@item
When displaying any @code{MySQL AB} logo except the @emph{Conditional Use}
logos mentioned previously on web sites or elsewhere.
@end itemize

Out of legal and commercial reasons we have to monitor the use of MySQL
trademarks on products, books, etc. We will usually require a fee for
displaying @code{MySQL AB} logos on commercial products, since we think
it is reasonable that some of the revenue is returned to fund further
development of the @code{MySQL} database.


@node MySQL AB Partnership Logos, Using MySQL word in presentations, When need MySQL logo permission, MySQL AB Logos and Trademarks
@subsubsection MySQL AB Partnership Logos

@c FIX: pictures: partnership logos - Bertrand?

@code{MySQL} partnership logos may only be used by companies and persons
having a written partnership agreement with @code{MySQL AB}. Partnerships
include certification as a @code{MySQL} trainer or consultant.
Please see @ref{Business Services Partnering,,Partnering}.


@node Using MySQL word in presentations, Using MySQL word in company and product names, MySQL AB Partnership Logos, MySQL AB Logos and Trademarks
@subsubsection Using the word @code{MySQL} in Printed Text or Presentations

@code{MySQL AB} welcomes references to the @code{MySQL} database, but
note that the word @code{MySQL} is a trademark of @code{MySQL AB}.
Because of this, you should append the trademark symbol (@code{TM}) to
the first or most prominent use of the word @code{MySQL} in a text and
where appropriate, state that @code{MySQL} is a trademark of
@code{MySQL AB}. Please refer to our trademark policy at
@uref{http://www.mysql.com/company/trademark.html} for details.


@node Using MySQL word in company and product names,  , Using MySQL word in presentations, MySQL AB Logos and Trademarks
@subsubsection Using the word @code{MySQL} in Company and Product Names

Use of the word @code{MySQL} in product or company names or in Internet
domain names is not allowed without written permission from @code{MySQL AB}.


@node MySQL 4.0 In A Nutshell, MySQL Information Sources, Licensing and Support, Introduction
@section MySQL 4.0 In A Nutshell

Dateline: 16 October 2001, Uppsala, Sweden

Long promised by @code{MySQL AB} and long awaited by our users,
MySQL Server 4.0 is now available in alpha version for download from
@uref{http://www.mysql.com/} and our mirrors.

Main new features of MySQL Server 4.0 are geared toward our existing
business and community users, enhancing the MySQL database software
as the solution for mission-critical, heavy-load database systems.
Other new features target the users of embedded databases.

@menu
* Nutshell Stepwise Rollout::   Stepwise Rollout
* Nutshell Ready for Immediate Development Use::  Ready for Immediate Development Use
* Nutshell Embedded MySQL::     Embedded MySQL
* Nutshell Other features::     Other Features Available From MySQL 4.0.0
* Nutshell Future features::    Future MySQL 4.0 Features
* Nutshell 4.1 development release::  MySQL 4.1, The Following Development Release
@end menu


@node Nutshell Stepwise Rollout, Nutshell Ready for Immediate Development Use, MySQL 4.0 In A Nutshell, MySQL 4.0 In A Nutshell
@subsection Stepwise Rollout

The rollout of MySQL Server 4.0 will come in several steps, with
the first version labelled 4.0.0 already containing most of the
new features. Additional features will be incorporated into
MySQL 4.0.1, 4.0.2, and onward; very probably within a couple of months,
MySQL 4.0 will be labelled beta. Further new features will then
be added in MySQL 4.1, which is targeted for alpha release
in third quarter 2002.


@node Nutshell Ready for Immediate Development Use, Nutshell Embedded MySQL, Nutshell Stepwise Rollout, MySQL 4.0 In A Nutshell
@subsection Ready for Immediate Development Use

Users are not recommended to switch their production systems
to MySQL Server 4.0 until it is released in beta version.
However, even the initial release has passed our extensive
test suite without any errors on any of the platforms we test on.
Due to the large number of new features, we thus recommend
MySQL Server 4.0 even in alpha form for development use, with
the release schedule of MySQL Server 4.0 being such that it will
reach stable state before the deployment of user applications
now under development.


@node Nutshell Embedded MySQL, Nutshell Other features, Nutshell Ready for Immediate Development Use, MySQL 4.0 In A Nutshell
@subsection Embedded MySQL

@code{libmysqld} makes MySQL Server suitable for a vastly expanded realm of
applications. Using the embedded MySQL server library, one can
embed MySQL Server into various applications and electronics devices, where
the end user has no knowledge of there actually being an underlying
database. Embedded MySQL Server is ideal for use behind
the scenes in internet appliances, public kiosks, turnkey
hardware/software combination units, high performance internet
servers, self-contained databases distributed on CD-ROM, etc.

Many users of @code{libmysqld} will benefit from the MySQL
@emph{Dual Licensing}. For those not wishing to be bound by the GPL,
the software is also made available under a commercial license.
The embedded MySQL library uses the same interface as the normal
client library, so it is convenient and easy to use. @xref{libmysqld}.


@node Nutshell Other features, Nutshell Future features, Nutshell Embedded MySQL, MySQL 4.0 In A Nutshell
@subsection Other Features Available From MySQL 4.0.0

@itemize @bullet
@item
Version 4.0 further increases @emph{the speed of MySQL Server}
in a number of areas,
such as bulk @code{INSERT}s, searching on packed indexes, creation of
@code{FULLTEXT} indexes, as well as @code{COUNT(DISTINCT)}.

@item
The table handler @code{InnoDB} is now offered as a feature of the
standard MySQL server, including full support for @code{transactions}
and @code{row-level locking}.

@item
MySQL Server 4.0 will support secure traffic between the client and the server,
greatly increasing security against malicious intrusion and unauthorised
access. Web applications being a cornerstone of MySQL use, web developers
have been able to use @code{Secure Socket Layer} (@code{SSL}) to secure the
traffic between the the end user browser and the web application, be it
written in PHP, Perl, ASP or using any other web development tool. However,
the traffic between the development tool and the @code{mysqld} server
process has been protected only by virtue of them being processes
residing on computers within the same firewall. In MySQL Server 4.0,
the @code{mysqld} server daemon process can itself use @code{SSL},
thus enabling secure traffic to MySQL databases from, say, a Windows
application residing outside the firewall.

@item
Our German, Austrian, and Swiss users will note that we have a new character
set, @code{latin_de}, which corrects the @emph{German sorting order},
placing German umlauts in the same order as German telephone books.

@item
Features to simplify migration from other database systems to MySQL
Server include @code{TRUNCATE TABLE} (like in Oracle) and @code{IDENTITY} as a
synonym for automatically incremented keys (like in Sybase). Many users will
also be happy to learn that MySQL Server now supports the @code{UNION} statement,
a long-awaited standard SQL feature.

@item
In the process of building features for new users, we have not forgotten
requests by the community of loyal users.
We have multi-table @code{DELETE} statements.
By adding support for @code{symbolic linking} to @code{MyISAM} on the table
level (and not just the database level as before), as well as by enabling symlink
handling by default on Windows, we hope to show that we take enhancement
requests seriously.
Functions like @code{SQL_CALC_FOUND_ROWS} and @code{FOUND_ROWS()} make it
possible to know how many rows a query would have returned without a
@code{LIMIT} clause.
@end itemize


@node Nutshell Future features, Nutshell 4.1 development release, Nutshell Other features, MySQL 4.0 In A Nutshell
@subsection Future MySQL 4.0 Features

For the upcoming MySQL Server 4.0 releases (4.0.1, 4.0.2, and onward),
expect the following features now still under development:

@itemize @bullet
@item
Mission-critical, heavy-load users of MySQL Server will appreciate
the additions to our replication system and our online hot backup.
Later versions of 4.0 will include @code{fail-safe replication};
already existing in 4.0.0, the @code{LOAD DATA FROM MASTER} command
will soon automate slave setup. The @code{online backup} will make
it easy to add a new replication slave without taking down
the master, and have a very low performance penalty on
update-heavy systems.

@item
A convenience feature for Database Administrators is that
@code{mysqld} parameters (startup options) can soon be set without
taking down the servers.

@item
The new @code{FULLTEXT} search properties of MySQL Server 4.0 enable the
use of @code{FULLTEXT} indexing of large text masses with both binary and
natural-language searching logic. Users can customise minimal word
length and define their own stop word lists in any human language,
enabling a new set of applications to be built on MySQL Server.

@item
Many read-heavy applications will benefit from
further increased speed through the rewritten @code{key cache}.

@item
Many developers will also be happy to see the @code{MySQL command help}
in the client.
@end itemize


@node Nutshell 4.1 development release,  , Nutshell Future features, MySQL 4.0 In A Nutshell
@subsection MySQL 4.1, The Following Development Release

Internally, through a new .frm file format for table definitions,
MySQL Server 4.0 lays the foundation for the new features of
MySQL Server 4.1 and onward,
such as @code{nested subqueries}, @code{stored procedures}, and
@code{foreign key integrity rules}, which form the top of the
wish list for many of our customers. Along with those, we will
also include simpler additions, such as
multi-table @code{UPDATE} statements.

After those additions, critics of the MySQL Database Server have
to be more imaginative
than ever in pointing out deficiencies in the MySQL Database
Management System. For long already known for its stability,
speed, and ease of use, MySQL Server will then match the requirement
checklist of very demanding buyers.


@node MySQL Information Sources, Compatibility, MySQL 4.0 In A Nutshell, Introduction
@section MySQL Information Sources

@menu
* Portals::                     MySQL Portals
* Questions::                   MySQL Mailing Lists
@end menu


@node Portals, Questions, MySQL Information Sources, MySQL Information Sources
@subsection MySQL Portals

@cindex MySQL Portals
@cindex manuals, about MySQL
@cindex books, about MySQL
@cindex MySQL Testimonials
@cindex users, of MySQL
@cindex news sites
@cindex search engines, web
@cindex web search engines
@cindex online magazines
@cindex magazines, online
@cindex web sites
@cindex services
@cindex PHP, web sites
@cindex consultants, list of
@cindex web pages, miscellaneous
@cindex @code{Contrib} directory
@cindex URLs to MySQL information
@cindex MySQL related information URLs

The @code{MySQL Portals} (@uref{http://www.mysql.com/portal/})
represent the ultimate resource to find @code{MySQL AB Partners},
as well as books, or other @code{MySQL}-related solutions that you
may be looking for. Items are categorised and rated in order to
make it easy for you to locate information.

By registering as a user, you will have the ability to comment on and
rate items presented in portals. You will also receive relevant
newsletters according to your user profile that you may update at
any time.

Some of the current @code{MySQL Portal} categories include:

@table @strong
@item Partners
Find @code{MySQL AB} partners worldwide.

@item Books
Comment on, vote for, and buy books related to @code{MySQL}.

@item Development
Various links to different sites that are using @code{MySQL Server}
for different purposes, with a description of each site.
This information can give you an idea of who uses the @code{MySQL}
database software and how @code{MySQL Server} can fulfill
requirements.

Let us know about @emph{your} site or success story, too!
Visit @uref{http://www.mysql.com/feedback/testimonial.php}.

@item Software
Find, buy, and download several applications and wrappers that make
use of the @code{MySQL} server.

@item Distributions
From here you can find the various Linux distributions and other
software packages that contain the @code{MySQL} software.

@item Service Providers
Companies providing @code{MySQL}-related services.
@end table


@node Questions,  , Portals, MySQL Information Sources
@subsection MySQL Mailing Lists

@cindex reporting, errors
@cindex errors, reporting
@cindex MySQL mailing lists

@menu
* Mailing-list::                The MySQL Mailing Lists
* Asking questions::            Asking Questions or Reporting Bugs
* Bug reports::                 How to Report Bugs or Problems
* Answering questions::         Guidelines for Answering Questions on the Mailing List
@end menu

This section introduces you to the MySQL mailing lists, and gives
some guidelines as to how to use them. By subscribing to a mailing
list, you will receive as e-mail messages all other postings on the list,
and you will be able to send in your own questions and answers.


@node Mailing-list, Asking questions, Questions, Questions
@subsubsection The MySQL Mailing Lists

@cindex mailing lists
@cindex e-mail lists

To subscribe to the main MySQL mailing list, send a message to the
electronic mail address @email{mysql-subscribe@@lists.mysql.com}.

To unsubscribe from the main MySQL mailing list, send a message to
the electronic mail address @email{mysql-unsubscribe@@lists.mysql.com}.

Only the address to which you send your messages is significant.  The
subject line and the body of the message are ignored.

@c the last two addresses in this paragraph are not @email because they
@c shouldn't be live links.
If your reply address is not valid, you can specify your address
explicitly, by adding a hyphen to the subscribe or unsubscribe command
word, followed by your address with the @samp{@@} character in your
address replaced by a @samp{=}.  For example, to subscribe
@code{your_name@@host.domain}, send a message to
@code{mysql-subscribe-your_name=host.domain@@lists.mysql.com}.

Mail to @email{mysql-subscribe@@lists.mysql.com} or
@email{mysql-unsubscribe@@lists.mysql.com} is handled automatically by the
ezmlm mailing list processor.  Information about ezmlm is available at
the ezmlm web site (@uref{http://www.ezmlm.org/}).

To post a message to the list itself, send your message to
@code{mysql@@lists.mysql.com}.  However, please @strong{do not} send mail about
subscribing or unsubscribing to @email{mysql@@lists.mysql.com} because any
mail sent to that address is distributed automatically to thousands of other
users.

Your local site may have many subscribers to @email{mysql@@lists.mysql.com}.
If so, it may have a local mailing list, so messages sent from
@code{lists.mysql.com} to your site are propagated to the local list.  In such
cases, please contact your system administrator to be added to or dropped
from the local MySQL list.

If you wish to have traffic for a mailing list go to a separate mailbox in
your mail program, set up a filter based on the message headers.  You can
use either the @code{List-ID:} or @code{Delivered-To:} headers to identify
list messages.

The following MySQL mailing lists exist:

@table @code
@item @email{announce-subscribe@@lists.mysql.com} announce
This is for announcement of new versions of MySQL and related
programs.  This is a low-volume list all MySQL users should
subscribe to.

@item @email{mysql-subscribe@@lists.mysql.com} mysql
The main list for general MySQL discussion.  Please note that some
topics are better discussed on the more-specialised lists.  If you post to the
wrong list, you may not get an answer!

@item @email{mysql-digest-subscribe@@lists.mysql.com} mysql-digest
The @code{mysql} list in digest form.  That means you get all individual
messages, sent as one large mail message once a day.

@item @email{bugs-subscribe@@lists.mysql.com} bugs
On this list you should only post a full, repeatable bug report using
the @code{mysqlbug} script (if you are running on Windows, you should
include a description of the operating system and the MySQL version).
Preferably, you should test the problem using the latest stable or development
version of MySQL Server before posting!  Anyone should be able to repeat the
bug by just using @code{mysql test < script} on the included test case.  All
bugs posted on this list will be corrected or documented in the next
MySQL release!  If only small code changes are needed, we
will also post a patch that fixes the problem.

@item @email{bugs-digest-subscribe@@lists.mysql.com} bugs-digest
The @code{bugs} list in digest form.

@item @email{internals-subscribe@@lists.mysql.com} internals
A list for people who work on the MySQL code.  On this list one
can also discuss MySQL development and post patches.

@item @email{internals-digest-subscribe@@lists.mysql.com} internals-digest
A digest version of the @code{internals} list.

@item @email{java-subscribe@@lists.mysql.com} java
Discussion about the MySQL server and Java.  Mostly about the JDBC drivers.

@item @email{java-digest-subscribe@@lists.mysql.com} java-digest
A digest version of the @code{java} list.

@item @email{win32-subscribe@@lists.mysql.com} win32
All things concerning the MySQL software on Microsoft operating systems
such as Windows 9x/Me/NT/2000/XP.

@item @email{win32-digest-subscribe@@lists.mysql.com} win32-digest
A digest version of the @code{win32} list.

@item @email{myodbc-subscribe@@lists.mysql.com} myodbc
All things about connecting to the MySQL server with ODBC.

@item @email{myodbc-digest-subscribe@@lists.mysql.com} myodbc-digest
A digest version of the @code{myodbc} list.

@item @email{mycc-subscribe@@lists.mysql.com} mycc
All things about the MySQL @code{MyCC} graphical client.

@item @email{mycc-digest-subscribe@@lists.mysql.com} mycc-digest
A digest version of the @code{mycc} list.

@item @email{plusplus-subscribe@@lists.mysql.com} plusplus
All things concerning programming with the C++ API to MySQL.

@item @email{plusplus-digest-subscribe@@lists.mysql.com} plusplus-digest
A digest version of the @code{plusplus} list.

@item @email{msql-mysql-modules-subscribe@@lists.mysql.com} msql-mysql-modules
A list about the Perl support for MySQL with msql-mysql-modules.

@item @email{msql-mysql-modules-digest-subscribe@@lists.mysql.com} msql-mysql-modules-digest
A digest version of the @code{msql-mysql-modules} list.
@end table

You subscribe or unsubscribe to all lists in the same way as described
previously.  In your subscribe or unsubscribe message, just put the appropriate
mailing list name rather than @code{mysql}.  For example, to subscribe to or
unsubscribe from the @code{myodbc} list, send a message to
@email{myodbc-subscribe@@lists.mysql.com} or
@email{myodbc-unsubscribe@@lists.mysql.com}.

If you can't get an answer for your questions from the mailing list, one
option is to pay for support from MySQL AB, which will put you
in direct contact with MySQL developers. @xref{Support}.

The following table shows some MySQL mailing in languages other than
English.  Note that these are not operated by MySQL AB, so we can't
guarantee the quality on these.

@table @code
@item @email{mysql-france-subscribe@@yahoogroups.com} A French mailing list
@item @email{list@@tinc.net} A Korean mailing list
E-mail @code{subscribe mysql your@@e-mail.address} to this list.
@item @email{mysql-de-request@@lists.4t2.com} A German mailing list
E-mail @code{subscribe mysql-de your@@e-mail.address} to this list.
You can find information about this mailing list at
@uref{http://www.4t2.com/mysql/}.
@item @email{mysql-br-request@@listas.linkway.com.br} A Portugese mailing list
E-mail @code{subscribe mysql-br your@@e-mail.address} to this list.
@item @email{mysql-alta@@elistas.net} A Spanish mailing list
E-mail @code{subscribe mysql your@@e-mail.address} to this list.
@end table


@node Asking questions, Bug reports, Mailing-list, Questions
@subsubsection Asking Questions or Reporting Bugs

@cindex net etiquette
@cindex mailing lists, archive location
@cindex searching, MySQL web pages

Before posting a bug report or question, please do the following:

@itemize @bullet
@item
Start by searching the MySQL online manual at:
@*
@uref{http://www.mysql.com/doc/}
@*
We try to keep the manual up to date by updating it frequently with
solutions to newly found problems!

@item
Search the MySQL mailing list archives:
@*
@uref{http://lists.mysql.com/}
@*
@item
You can also use @uref{http://www.mysql.com/search.html} to search all the
web pages (including the manual) that are located at
@uref{http://www.mysql.com/}.
@end itemize

If you can't find an answer in the manual or the archives, check with your
local MySQL expert.  If you still can't find an answer to your
question, go ahead and read the next section about how to send mail to
@email{mysql@@lists.mysql.com}.


@node Bug reports, Answering questions, Asking questions, Questions
@subsubsection How to Report Bugs or Problems

@cindex bugs, reporting
@cindex reporting, bugs
@cindex problems, reporting
@cindex errors, reporting
@cindex @code{mysqlbug} script
@cindex creating, bug reports
@cindex scripts, @code{mysqlbug}

Writing a good bug report takes patience, but doing it right the first
time saves time for us and for you.  A good bug report containing a full
test case for the bug will make it very likely that we will fix it in
the next release.  This section will help you write your report correctly
so that you don't waste your time doing things that may not help us much
or at all.

We encourage everyone to use the @code{mysqlbug} script to generate a bug
report (or a report about any problem), if possible.  @code{mysqlbug} can be
found in the @file{scripts} directory in the source distribution, or for a
binary distribution, in the @file{bin} directory under your MySQL
installation directory.  If you are unable to use @code{mysqlbug}, you should
still include all the necessary information listed in this section.

The @code{mysqlbug} script helps you generate a report by determining much
of the following information automatically, but if something important is
missing, please include it with your message!  Please read this section
carefully and make sure that all the information described here is included
in your report.

@cindex bug reports, e-mail address
The normal place to report bugs and problems is
@email{mysql@@lists.mysql.com}.  If you can make a test case that clearly
demonstrates the bug, you should post it to the @email{bugs@@lists.mysql.com}
list.  Note that on this list you should only post a full, repeatable bug
report using the @code{mysqlbug} script.  If you are running on Windows,
you should include a description of the operating system and the
MySQL version.  Preferably, you should test the problem using
the latest stable or development version of MySQL Server before
posting!  Anyone should be able to repeat the bug by just using
``@code{mysql test < script}'' on the included test case or run the
shell or Perl script that is included in the bug report.  All bugs
posted on the @code{bugs} list will be corrected or documented in the next
MySQL release!  If only small code changes are needed
to correct this problem, we will also post a patch that fixes the
problem.

If you have found a sensitive security bug in MySQL, you should send an
e-mail to @email{security@@mysql.com}.

Remember that it is possible to respond to a message containing too much
information, but not to one containing too little.  Often people omit facts
because they think they know the cause of a problem and assume that some
details don't matter.  A good principle is: if you are in doubt about stating
something, state it!  It is a thousand times faster and less troublesome to
write a couple of lines more in your report than to be forced to ask again
and wait for the answer because you didn't include enough information the
first time.

The most common errors are that people don't indicate the version number of
the MySQL distribution they are using, or don't indicate what
platform they have the MySQL server installed on (including the platform
version number).  This is highly relevant information, and in 99 cases out of
100 the bug report is useless without it!  Very often we get questions like,
``Why doesn't this work for me?'' Then we find that the feature
requested wasn't implemented in that MySQL version, or that a bug
described in a report has been fixed already in newer MySQL
versions.  Sometimes the error is platform-dependent; in such cases, it is
next to impossible to fix anything without knowing the operating system and
the version number of the platform.

Remember also to provide information about your compiler, if it is related to
the problem.  Often people find bugs in compilers and think the problem is
MySQL-related.  Most compilers are under development all the time and
become better version by version.  To determine whether your
problem depends on your compiler, we need to know what compiler is used.
Note that every compiling problem should be regarded as a bug report and
reported accordingly.

It is most helpful when a good description of the problem is included in the
bug report.  That is, a good example of all the things you did that led to
the problem and the problem itself exactly described.  The best reports are
those that include a full example showing how to reproduce the bug or
problem.  @xref{Reproduceable test case}.

If a program produces an error message, it is very important to include the
message in your report!  If we try to search for something from the archives
using programs, it is better that the error message reported exactly matches
the one that the program produces.  (Even the case should be observed!)
You should never try to remember what the error message was; instead, copy
and paste the entire message into your report!

If you have a problem with MyODBC, you should try to generate a MyODBC
trace file.  @xref{MyODBC bug report}.

Please remember that many of the people who will read your report will
do so using an 80-column display.  When generating reports or examples
using the @code{mysql} command-line tool, you should therefore use
the @code{--vertical} option (or the @code{\G} statement terminator)
for output that would exceed the available width for such a display
(for example, with the @code{EXPLAIN SELECT} statement; see the
example later in this section).

@cindex bug reports, criteria for
Please include the following information in your report:

@itemize @bullet
@item
The version number of the MySQL distribution you are using (for
example, MySQL Version 3.22.22). You can find out which version you
are running by executing @code{mysqladmin version}.  @code{mysqladmin} can be
found in the @file{bin} directory under your MySQL installation
directory.

@item
The manufacturer and model of the machine you are working on.

@item
The operating system name and version.  For most operating systems, you can
get this information by executing the Unix command @code{uname -a}.

@item
Sometimes the amount of memory (real and virtual) is relevant. If in doubt,
include these values.

@item
If you are using a source distribution of the MySQL software, the name and
version number of the compiler used is needed.  If you have a binary
distribution, the distribution name is needed.

@item
If the problem occurs during compilation, include the exact error
message(s) and also a few lines of context around the offending code in the
file where the error occurred.

@item
If @code{mysqld} died, you should also report the query that crashed
@code{mysqld}.  You can usually find this out by running @code{mysqld} with
logging enabled.  @xref{Using log files}.

@item
If any database table is related to the problem, include the output from
@code{mysqldump --no-data db_name tbl_name1 tbl_name2 ...}.  This is very easy
to do and is a powerful way to get information about any table in a database
that will help us create a situation matching the one you have.

@item
For speed-related bugs or problems with @code{SELECT} statements, you should
always include the output of @code{EXPLAIN SELECT ...}, and at least the
number of rows that the @code{SELECT} statement produces.  The more
information you give about your situation, the more likely it is that someone
can help you!  For example, the following is an example of a very good bug
report (it should of course be posted with the @code{mysqlbug} script):

Example run using the @code{mysql} command-line tool (note the use of the
@code{\G} statement terminator for statements whose output width would
otherwise exceed that of an 80-column display device):

@example
mysql> SHOW VARIABLES;
mysql> SHOW COLUMNS FROM ...\G
       <output from SHOW COLUMNS>
mysql> EXPLAIN SELECT ...\G
       <output from EXPLAIN>
mysql> FLUSH STATUS;
mysql> SELECT ...;
       <A short version of the output from SELECT,
       including the time taken to run the query>
mysql> SHOW STATUS;
       <output from SHOW STATUS>
@end example

@item
If a bug or problem occurs while running @code{mysqld}, try to provide an
input script that will reproduce the anomaly.  This script should include any
necessary source files.  The more closely the script can reproduce your
situation, the better.  If you can make a reproduceable test case, you should
post this to @email{bugs@@lists.mysql.com} for a high-priority treatment!

If you can't provide a script, you should at least include the output
from @code{mysqladmin variables extended-status processlist} in your mail to
provide some information of how your system is performing!

@item
If you can't produce a test case in a few rows, or if the test table
is too big to be mailed to the mailing list (more than 10 rows), you should
dump your tables using @code{mysqldump} and create a @file{README} file
that describes your problem.

Create a compressed archive of your files using
@code{tar} and @code{gzip} or @code{zip}, and use @code{ftp} to transfer the
archive to @uref{ftp://support.mysql.com/pub/mysql/secret/}.  Then send a
short description of the problem to @email{bugs@@lists.mysql.com}.

@item
If you think that the MySQL server produces a strange result from a query,
include not only the result, but also your opinion of what the result
should be, and an account describing the basis for your opinion.

@item
When giving an example of the problem, it's better to use the variable names,
table names, etc., that exist in your actual situation than to come up with
new names.  The problem could be related to the name of a variable or table!
These cases are rare, perhaps, but it is better to be safe than sorry.
After all, it should be easier for you to provide an example that uses your
actual situation, and it is by all means better for us.  In case you have data
you don't want to show to others, you can use @code{ftp} to transfer it to
@uref{ftp://support.mysql.com/pub/mysql/secret/}.  If the data is really top
secret and you don't want to show it even to us, then go ahead and provide
an example using other names, but please regard this as the last choice.

@item
Include all the options given to the relevant programs, if possible.  For
example, indicate the options that you use when you start the @code{mysqld}
daemon and that you use to run any MySQL client programs.  The
options to programs like @code{mysqld} and @code{mysql}, and to the
@code{configure} script, are often keys to answers and are very relevant!
It is never a bad idea to include them anyway!  If you use any modules, such
as Perl or PHP, please include the version number(s) of those as well.

@item
If your question is related to the privilege system, please include the
output of @code{mysqlaccess}, the output of @code{mysqladmin reload}, and all
the error messages you get when trying to connect!  When you test your
privileges, you should first run @code{mysqlaccess}.  After this, execute
@code{mysqladmin reload version} and try to connect with the program that
gives you trouble.  @code{mysqlaccess} can be found in the @file{bin}
directory under your MySQL installation directory.

@item
If you have a patch for a bug, that is good. But don't assume the patch is
all we need, or that we will use it, if you don't provide some necessary
information such as test cases showing the bug that your patch fixes.  We
might find problems with your patch or we might not understand it at all; if
so, we can't use it.

If we can't verify exactly what the patch is meant for, we won't use it.
Test cases will help us here.  Show that the patch will handle all the
situations that may occur.  If we find a borderline case (even a rare one)
where the patch won't work, it may be useless.

@item
Guesses about what the bug is, why it occurs, or what it depends on
are usually wrong.  Even the MySQL team can't guess such things
without first using a debugger to determine the real cause of a bug.

@item
Indicate in your mail message that you have checked the reference manual
and mail archive so that others know you have tried to solve the
problem yourself.

@item
If you get a @code{parse error}, please check your syntax closely!  If
you can't find something wrong with it, it's extremely likely that your
current version of MySQL Server doesn't support the query you are
using.  If you are using the current version and the manual at
@uref{http://www.mysql.com/doc/} doesn't cover the
syntax you are using, MySQL Server doesn't support your query.  In this
case, your only options are to implement the syntax yourself or e-mail
@email{licensing@@mysql.com} and ask for an offer to implement it!

If the manual covers the syntax you are using, but you have an older version
of MySQL Server, you should check the MySQL change history to see
when the syntax was implemented.  In this case, you have the option of
upgrading to a newer version of MySQL Server.  @xref{News}.

@item
If you have a problem such that your data appears corrupt or you get
errors when you access some particular table, you should first check and then
try repairing your tables with @code{myisamchk} or @code{CHECK TABLE} and
@code{REPAIR TABLE}.  @xref{MySQL Database Administration}.

@item
If you often get corrupted tables you should try to find out when and why this
happens.  In this case, the @file{mysql-data-directory/'hostname'.err} file
may contain some information about what happened.  @xref{Error log}.  Please
include any relevant information from this file in your bug report.  Normally
@code{mysqld} should @strong{never} crash a table if nothing killed it in the
middle of an update!  If you can find the cause of @code{mysqld} dying,
it's much easier for us to provide you with a fix for the problem.
@xref{What is crashing}.

@item
If possible, download and install the most recent version of MySQL Server
and check whether it solves your problem.  All versions of
the MySQL software are thoroughly tested and should work without problems.
We believe in making everything as backward-compatible as possible,
and you  should be able to switch MySQL versions without any hassle.
@xref{Which version}.
@end itemize

@cindex technical support, mailing address
@cindex support, mailing address
@cindex customer support, mailing address
@cindex mailing address, for customer support
If you are a support customer, please cross-post the bug report to
@email{mysql-support@@mysql.com} for higher-priority treatment, as well as to
the appropriate mailing list to see if someone else has experienced (and
perhaps solved) the problem.

For information on reporting bugs in @code{MyODBC}, see @ref{ODBC Problems}.

For solutions to some common problems, see @ref{Problems}.

When answers are sent to you individually and not to the mailing list,
it is considered good etiquette to summarise the answers and send the
summary to the mailing list so that others may have the benefit of
responses you received that helped you solve your problem!


@node Answering questions,  , Bug reports, Questions
@subsubsection Guidelines for Answering Questions on the Mailing List

@cindex net etiquette
@cindex questions, answering
@cindex answering questions, etiquette
@cindex mailing lists, guidelines

If you consider your answer to have broad interest, you may want to post it
to the mailing list instead of replying directly to the individual who
asked.  Try to make your answer general enough that people other than the
original poster may benefit from it.  When you post to the list, please make
sure that your answer is not a duplication of a previous answer.

Try to summarise the essential part of the question in your reply; don't feel
obliged to quote the entire original message.

Please don't post mail messages from your browser with HTML mode turned on!
Many users don't read mail with a browser!


@node Compatibility, TODO, MySQL Information Sources, Introduction
@section How Standards-compatible Is MySQL?

@cindex compatibility, with ANSI SQL
@cindex standards compatibility
@cindex extensions, to ANSI SQL
@cindex ANSI SQL92, extensions to

This section describes how MySQL relates to the ANSI SQL standards.
MySQL Server has many extensions to the ANSI SQL standards, and here you
will find out what they are and how to use them.  You will also find
information about functionality missing from MySQL Server, and how to work
around some differences.

Our goal is to not, without a very good reason, restrict MySQL Server usability
for any usage.  Even if we don't have the resources to do development
for every possible use, we are always willing to help and offer
suggestions to people who are trying to use MySQL Server in new territories.

One of our main goals with the product is to continue to work toward
ANSI 99 compliancy, but without sacrificing speed or reliability.
We are not afraid to add extensions to SQL or support for non-SQL
features if this greatly increases the usability of MySQL Server for a big
part of our users.  (The new @code{HANDLER} interface in MySQL Server 4.0
is an example of this strategy. @xref{HANDLER, , @code{HANDLER}}.)

We will continue to support transactional and non-transactional
databases to satisfy both heavy web/logging usage and mission-critical
24/7 usage.

MySQL Server was designed from the start to work with medium size databases
(10-100 million rows, or about 100 MB per table) on small computer
systems.  We will continue to extend MySQL Server to work even better
with terabyte-size databases, as well as to make it possible
to compile a reduced MySQL version that is more suitable for hand-held
devices and embedded usage.  The compact design of the MySQL server makes both
of these directions possible without any conflicts in the source tree.

We are currently not targeting realtime support or clustered databases
(even if you can already do a lot of things with our replication
services).

We don't believe that one should have native XML support in the
database, but will instead add the XML support our users request from
us on the client side.  We think it's better to keep the main server
code as ``lean and clean'' as possible and instead develop libraries to
deal with the complexity on the client side.  This is part of the strategy
mentioned previously of not sacrificing speed or reliability in the
server.

@menu
* Standards::                   What Standards Does MySQL Follow?
* ANSI mode::                   Running MySQL in ANSI Mode
* Extensions to ANSI::          MySQL Extensions to ANSI SQL92
* Differences from ANSI::       MySQL Differences Compared to ANSI SQL92
* Bugs::                        Known Errors and Design Deficiencies in MySQL
@end menu


@node Standards, ANSI mode, Compatibility, Compatibility
@subsection What Standards Does MySQL Follow?

Entry-level SQL92. ODBC levels 0-3.51.

We are aiming toward supporting the full ANSI SQL99 standard,
but without concessions to speed and quality of the code.


@node ANSI mode, Extensions to ANSI, Standards, Compatibility
@subsection Running MySQL in ANSI Mode

@cindex running, ANSI mode
@cindex ANSI mode, running

If you start @code{mysqld} with the @code{--ansi} option, the following
behavior of MySQL Server changes:

@itemize @bullet
@item
@code{||} is string concatenation instead of @code{OR}.

@item
You can have any number of spaces between a function name and the @samp{(}.
This forces all function names to be treated as reserved words.

@item
@samp{"} will be an identifier quote character (like the MySQL Server
@samp{`} quote character) and not a string quote character.

@item
@code{REAL} will be a synonym for @code{FLOAT} instead of a synonym for
@code{DOUBLE}.

@item
The default transaction isolation level is @code{SERIALIZABLE}.
@xref{SET TRANSACTION}.
@end itemize

This is the same as using
@code{--sql-mode=REAL_AS_FLOAT,PIPES_AS_CONCAT,ANSI_QUOTES,
IGNORE_SPACE,SERIALIZE,ONLY_FULL_GROUP_BY}.


@node Extensions to ANSI, Differences from ANSI, ANSI mode, Compatibility
@subsection MySQL Extensions to ANSI SQL92

@cindex hints
MySQL Server includes some extensions that you probably will not find in
other SQL databases.  Be warned that if you use them, your code will not be
portable to other SQL servers.  In some cases, you can write code that
includes MySQL extensions, but is still portable, by using comments
of the form @code{/*! ... */}.  In this case, MySQL Server will parse and
execute the code within the comment as it would any other MySQL
statement, but other SQL servers will ignore the extensions.  For example:

@example
SELECT /*! STRAIGHT_JOIN */ col_name FROM table1,table2 WHERE ...
@end example

If you add a version number after the @code{'!'}, the syntax will be
executed only if the MySQL version is equal to or newer than the used
version number:

@example
CREATE /*!32302 TEMPORARY */ TABLE (a int);
@end example

This means that if you have Version 3.23.02 or newer, MySQL
Server will use the @code{TEMPORARY} keyword.

The following is a list of MySQL extensions:

@itemize @bullet
@item
The field types @code{MEDIUMINT}, @code{SET}, @code{ENUM}, and the
different @code{BLOB} and @code{TEXT} types.

@item
The field attributes @code{AUTO_INCREMENT}, @code{BINARY}, @code{NULL},
@code{UNSIGNED}, and @code{ZEROFILL}.

@item
All string comparisons are case-insensitive by default, with sort
ordering determined by the current character set (ISO-8859-1 Latin1 by
default).  If you don't like this, you should declare your columns with
the @code{BINARY} attribute or use the @code{BINARY} cast, which causes
comparisons to be done according to the ASCII order used on the
MySQL server host.

@item
MySQL Server maps each database to a directory under the MySQL
data directory, and tables within a database to filenames in the database
directory.

This has a few implications:

@cindex database names, case sensitivity
@cindex table names, case sensitivity
@cindex case sensitivity, of database names
@cindex case sensitivity, of table names

@itemize @minus

@item
Database names and table names are case-sensitive in MySQL Server on
operating systems that have case-sensitive filenames (like most Unix
systems). @xref{Name case sensitivity}.

@item
Database, table, index, column, or alias names may begin with a digit
(but may not consist solely of digits).

@item
You can use standard system commands to back up, rename, move, delete, and copy
tables.  For example, to rename a table, rename the @file{.MYD}, @file{.MYI},
and @file{.frm} files to which the table corresponds.
@end itemize

@item
In SQL statements, you can access tables from different databases
with the @code{db_name.tbl_name} syntax.  Some SQL servers provide
the same functionality but call this @code{User space}.
MySQL Server doesn't support tablespaces as in:
@code{create table ralph.my_table...IN my_tablespace}.

@item
@code{LIKE} is allowed on numeric columns.

@item
Use of @code{INTO OUTFILE} and @code{STRAIGHT_JOIN} in a @code{SELECT}
statement. @xref{SELECT, , @code{SELECT}}.

@item
The @code{SQL_SMALL_RESULT} option in a @code{SELECT} statement.

@item
@code{EXPLAIN SELECT} to get a description on how tables are joined.

@item
Use of index names, indexes on a prefix of a field, and use of
@code{INDEX} or @code{KEY} in a @code{CREATE TABLE}
statement. @xref{CREATE TABLE, , @code{CREATE TABLE}}.

@item
Use of @code{TEMPORARY} or @code{IF NOT EXISTS} with @code{CREATE TABLE}.

@item
Use of @code{COUNT(DISTINCT list)} where @code{list} is more than one element.

@item
Use of @code{CHANGE col_name}, @code{DROP col_name}, or @code{DROP
INDEX}, @code{IGNORE} or @code{RENAME} in an @code{ALTER TABLE}
statement. @xref{ALTER TABLE, , @code{ALTER TABLE}}.

@item
Use of @code{RENAME TABLE}. @xref{RENAME TABLE, , @code{RENAME TABLE}}.

@item
Use of multiple @code{ADD}, @code{ALTER}, @code{DROP}, or @code{CHANGE}
clauses in an @code{ALTER TABLE} statement.

@item
Use of @code{DROP TABLE} with the keywords @code{IF EXISTS}.

@item
You can drop multiple tables with a single @code{DROP TABLE} statement.

@item
The @code{LIMIT} clause of the @code{DELETE} statement.

@item
The @code{DELAYED} clause of the @code{INSERT} and @code{REPLACE}
statements.

@item
The @code{LOW_PRIORITY} clause of the @code{INSERT}, @code{REPLACE},
@code{DELETE}, and @code{UPDATE} statements.

@cindex Oracle compatibility
@cindex compatibility, with Oracle
@item
Use of @code{LOAD DATA INFILE}. In many cases, this syntax is compatible with
Oracle's @code{LOAD DATA INFILE}. @xref{LOAD DATA, , @code{LOAD DATA}}.

@item
The @code{ANALYZE TABLE}, @code{CHECK TABLE}, @code{OPTIMIZE TABLE}, and
@code{REPAIR TABLE} statements.

@item
The @code{SHOW} statement.
@xref{SHOW, , @code{SHOW}}.

@item
Strings may be enclosed by either @samp{"} or @samp{'}, not just by @samp{'}.

@item
Use of the escape @samp{\} character.

@item
The @code{SET OPTION} statement. @xref{SET OPTION, , @code{SET OPTION}}.

@item
You don't need to name all selected columns in the @code{GROUP BY} part.
This gives better performance for some very specific, but quite normal
queries.
@xref{Group by functions}.

@item
One can specify @code{ASC} and @code{DESC} with @code{GROUP BY}.

@item
To make it easier for users who come from other SQL environments,
MySQL Server supports aliases for many functions. For example, all
string functions support both ANSI SQL syntax and ODBC syntax.

@item
MySQL Server understands the @code{||} and @code{&&} operators to mean
logical OR and AND, as in the C programming language.  In MySQL Server,
@code{||} and @code{OR} are synonyms, as are @code{&&} and @code{AND}.
Because of this nice syntax, MySQL Server doesn't support
the ANSI SQL @code{||} operator for string concatenation; use
@code{CONCAT()} instead. Because @code{CONCAT()} takes any number
of arguments, it's easy to convert use of the @code{||} operator to
MySQL Server.

@item
@code{CREATE DATABASE} or @code{DROP DATABASE}.
@xref{CREATE DATABASE, , @code{CREATE DATABASE}}.

@cindex PostgreSQL compatibility
@cindex compatibility, with PostgreSQL
@item
The @code{%} operator is a synonym for @code{MOD()}.  That is,
@code{N % M} is equivalent to @code{MOD(N,M)}.  @code{%} is supported
for C programmers and for compatibility with PostgreSQL.

@item
The @code{=}, @code{<>}, @code{<=} ,@code{<}, @code{>=},@code{>},
@code{<<}, @code{>>}, @code{<=>}, @code{AND}, @code{OR}, or @code{LIKE}
operators may be used in column comparisons to the left of the
@code{FROM} in @code{SELECT} statements.  For example:

@example
mysql> SELECT col1=1 AND col2=2 FROM tbl_name;
@end example

@item
The @code{LAST_INSERT_ID()} function.
@xref{mysql_insert_id, , @code{mysql_insert_id()}}.

@item
The @code{REGEXP} and @code{NOT REGEXP} extended regular expression
operators.

@item
@code{CONCAT()} or @code{CHAR()} with one argument or more than two
arguments.  (In MySQL Server, these functions can take any number of
arguments.)

@item The @code{BIT_COUNT()}, @code{CASE}, @code{ELT()},
@code{FROM_DAYS()}, @code{FORMAT()}, @code{IF()}, @code{PASSWORD()},
@code{ENCRYPT()}, @code{MD5()}, @code{ENCODE()}, @code{DECODE()},
@code{PERIOD_ADD()}, @code{PERIOD_DIFF()}, @code{TO_DAYS()}, or
@code{WEEKDAY()} functions.

@item
Use of @code{TRIM()} to trim substrings. ANSI SQL only supports removal
of single characters.

@item
The @code{GROUP BY} functions @code{STD()}, @code{BIT_OR()}, and
@code{BIT_AND()}.

@item
Use of @code{REPLACE} instead of @code{DELETE} + @code{INSERT}.
@xref{REPLACE, , @code{REPLACE}}.

@item
The @code{FLUSH}, @code{RESET} and @code{DO} statements.

@item
The ability to set variables in a statement with @code{:=}:
@example
SELECT @@a:=SUM(total),@@b=COUNT(*),@@a/@@b AS avg FROM test_table;
SELECT @@t1:=(@@t2:=1)+@@t3:=4,@@t1,@@t2,@@t3;
@end example

@end itemize


@node Differences from ANSI, Bugs, Extensions to ANSI, Compatibility
@subsection MySQL Differences Compared to ANSI SQL92

We try to make MySQL Server follow the ANSI SQL standard and the
ODBC SQL standard, but in some cases MySQL Server does things
differently:

@itemize @bullet
@item
For @code{VARCHAR} columns, trailing spaces are removed when the value is
stored. @xref{Bugs}.

@item
In some cases, @code{CHAR} columns are silently changed to @code{VARCHAR}
columns. @xref{Silent column changes}.

@item
Privileges for a table are not automatically revoked when you delete a
table. You must explicitly issue a @code{REVOKE} to revoke privileges for
a table. @xref{GRANT, , @code{GRANT}}.

@item
@code{NULL AND FALSE} will evaluate to @code{NULL} and not to @code{FALSE}.
This is because we don't think it's good to have to evaluate a lot of
extra conditions in this case.
@end itemize

@menu
* ANSI diff Sub-selects::       Sub-@code{SELECT}s
* ANSI diff SELECT INTO TABLE::  @code{SELECT INTO TABLE}
* ANSI diff Transactions::      Transactions and Atomic Operations
* ANSI diff Triggers::          Stored Procedures and Triggers
* ANSI diff Foreign Keys::      Foreign Keys
* ANSI diff Views::             Views
* ANSI diff comments::          @samp{--} as the Start of a Comment
@end menu

For a prioritised list indicating when new extensions will be added to
MySQL Server, you should consult the online MySQL TODO list at
@uref{http://www.mysql.com/documentation/manual.php?section=TODO}.
That is the latest version of the TODO list in this manual. @xref{TODO}.


@node ANSI diff Sub-selects, ANSI diff SELECT INTO TABLE, Differences from ANSI, Differences from ANSI
@subsubsection Sub-@code{SELECT}s

@cindex sub-selects

MySQL Server currently only supports nested queries of the form
@code{INSERT ... SELECT ...} and @code{REPLACE ... SELECT ...}.
You can, however, use the function @code{IN()} in other contexts.
Sub-selects are scheduled for implementation in Version 4.x.

Meanwhile, you can often rewrite the query without a sub-select:

@example
SELECT * FROM table1 WHERE id IN (SELECT id FROM table2);
@end example

This can be rewritten as:

@example
SELECT table1.* FROM table1,table2 WHERE table1.id=table2.id;
@end example

The queries:
@example
SELECT * FROM table1 WHERE id NOT IN (SELECT id FROM table2);
SELECT * FROM table1 WHERE NOT EXISTS (SELECT id FROM table2
                                       WHERE table1.id=table2.id);
@end example

Can be rewritten as:

@example
SELECT table1.* FROM table1 LEFT JOIN table2 ON table1.id=table2.id
                                       WHERE table2.id IS NULL;
@end example

For more complicated subqueries you can often create temporary tables
to hold the subquery.  In some cases, however, this option will not
work. The most frequently encountered of these cases arises with
@code{DELETE} statements, for which standard SQL does not support joins
(except in sub-selects).  For this situation there are two options
available until subqueries are supported by MySQL Server.

The first option is to use a procedural programming language (such as
Perl or PHP) to submit a @code{SELECT} query to obtain the primary keys
for the records to be deleted, and then use these values to construct
the @code{DELETE} statement (@code{DELETE FROM ... WHERE ... IN (key1,
key2, ...)}).

The second option is to use interactive SQL to construct a set of
@code{DELETE} statements automatically, using the MySQL
extension @code{CONCAT()} (in lieu of the standard @code{||} operator).
For example:

@example
SELECT CONCAT('DELETE FROM tab1 WHERE pkid = ', "'", tab1.pkid, "'", ';')
  FROM tab1, tab2
 WHERE tab1.col1 = tab2.col2;
@end example

You can place this query in a script file and redirect input from it to
the @code{mysql} command-line interpreter, piping its output back to a
second instance of the interpreter:

@example
shell> mysql --skip-column-names mydb < myscript.sql | mysql mydb
@end example

MySQL Server 4.0 supports multi-table deletes that can be used to
efficiently delete rows based on information from one table or even
from many tables at the same time.


@node ANSI diff SELECT INTO TABLE, ANSI diff Transactions, ANSI diff Sub-selects, Differences from ANSI
@subsubsection @code{SELECT INTO TABLE}

@findex SELECT INTO TABLE

MySQL Server doesn't yet support the Oracle SQL extension:
@code{SELECT ... INTO TABLE ...}.  MySQL Server supports instead the
ANSI SQL syntax @code{INSERT INTO ... SELECT ...}, which is basically
the same thing. @xref{INSERT SELECT}.

@example
INSERT INTO tblTemp2 (fldID) SELECT tblTemp1.fldOrder_ID
       FROM tblTemp1 WHERE tblTemp1.fldOrder_ID > 100;
@end example

Alternatively, you can use @code{SELECT INTO OUTFILE...} or @code{CREATE
TABLE ... SELECT}.


@node ANSI diff Transactions, ANSI diff Triggers, ANSI diff SELECT INTO TABLE, Differences from ANSI
@subsubsection Transactions and Atomic Operations

@findex COMMIT
@findex ROLLBACK
@cindex transactions, support
@cindex transaction-safe tables
@cindex tables, updating
@cindex updating, tables
@cindex @code{InnoDB} tables
@cindex @code{BDB} tables
@cindex @code{ACID}

MySQL Server supports transactions with the @code{InnoDB} and @code{BDB}
@code{Transactional table handlers}. @xref{Table types}.
@code{InnoDB} provides @code{ACID} compliancy.

However, the non-transactional table types in MySQL Server such as
@code{MyISAM} follow another paradigm for data integrity called
``@code{Atomic Operations}.'' Atomic operations often offer equal or
even better integrity with much better performance.
With MySQL Server supporting both paradigms, the user is able to decide if
he needs the speed of atomic operations or if he need to use
transactional features in his applications. This choice can be made
on a per-table basis.

How does one use the features of MySQL Server to maintain rigorous integrity
and how do these features compare with the transactional paradigm?

@enumerate
@item
In the transactional paradigm, if your applications are written in a
way that is dependent on the calling of @code{ROLLBACK} instead of
@code{COMMIT} in critical situations, transactions are more
convenient. Transactions also ensure that unfinished updates or
corrupting activities are not committed to the database; the server is
given the opportunity to do an automatic rollback and your database is
saved.

MySQL Server, in almost all cases, allows you to resolve potential problems
by including simple checks before updates and by running simple scripts
that check the databases for inconsistencies and automatically repair
or warn if such an inconsistency occurs. Note that just by using the
MySQL log or even adding one extra log, one can normally fix tables
perfectly with no data integrity loss.

@item
More often than not, fatal transactional updates can be rewritten to be
atomic. Generally speaking, all integrity problems that transactions
solve can be done with @code{LOCK TABLES} or atomic updates, ensuring
that you never will get an automatic abort from the database, which is
a common problem with transactional databases.

@item
Even a transactional system can lose data if the server goes down.
The difference between different systems lies in just how small the
time-lap is where they could lose data. No system is 100% secure, only
``secure enough.'' Even Oracle, reputed to be the safest of
transactional databases, is reported to sometimes lose data in such
situations.

To be safe with MySQL Server, whether using transactional tables or not, you
only need to have backups and have the update logging turned on. With
this you can recover from any situation that you could with any
other transactional database.  It is, of course, always good to have
backups, independent of which database you use.
@end enumerate

The transactional paradigm has its benefits and its drawbacks. Many
users and application developers depend on the ease with which they
can code around problems where an abort appears to be, or is necessary.
However, even if you are new to the atomic operations paradigm, or more
familiar with transactions, do consider the speed benefit that
non-transactional tables can offer on the order of three to five times
the speed of the fastest and most optimally tuned transactional tables.

In situations where integrity is of highest importance, MySQL Server offers
transaction-level or better reliability and integrity even for
non-transactional tables.
If you lock tables with @code{LOCK TABLES}, all updates will stall
until any integrity checks are made. If you only obtain a read lock
(as opposed to a write lock), reads and inserts are still allowed
to happen. The new inserted records will not be seen by any of the
clients that have a read lock until they release their read
locks. With @code{INSERT DELAYED} you can queue inserts into a local
queue, until the locks are released, without having the client wait
for the insert to complete. @xref{INSERT DELAYED}.

``Atomic,'' in the sense that we mean it, is nothing magical. It only
means that you can be sure that while each specific update is running,
no other user can interfere with it, and there will never be an
automatic rollback (which can happen with transactional tables if you
are not very careful).  MySQL Server also guarantees that there will not be
any dirty reads.

Following are some techniques for working with non-transactional tables:

@itemize @bullet
@item
Loops that need transactions normally can be coded with the help of
@code{LOCK TABLES}, and you don't need cursors when you can update
records on the fly.

@item
To avoid using @code{ROLLBACK}, you can use the following strategy:

@enumerate
@item
Use @code{LOCK TABLES ...} to lock all the tables you want to access.
@item
Test conditions.
@item
Update if everything is okay.
@item
Use @code{UNLOCK TABLES} to release your locks.
@end enumerate

This is usually a much faster method than using transactions with
possible @code{ROLLBACK}s, although not always. The only situation
this solution doesn't handle is when someone kills the threads in the
middle of an update. In this case, all locks will be released but some
of the updates may not have been executed.

@item
You can also use functions to update records in a single operation.
You can get a very efficient application by using the following
techniques:

@itemize @bullet
@item Modify fields relative to their current value.
@item Update only those fields that actually have changed.
@end itemize

For example, when we are doing updates to some customer information, we
update only the customer data that has changed and test only that none of
the changed data, or data that depends on the changed data, has changed
compared to the original row. The test for changed data is done with the
@code{WHERE} clause in the @code{UPDATE} statement. If the record wasn't
updated, we give the client a message: "Some of the data you have changed
has been changed by another user." Then we show the old row versus the new
row in a window, so the user can decide which version of the customer record
he should use.

This gives us something that is similar to column locking but is actually
even better because we only update some of the columns, using values that
are relative to their current values.  This means that typical @code{UPDATE}
statements look something like these:

@example
UPDATE tablename SET pay_back=pay_back+'relative change';

UPDATE customer
  SET
    customer_date='current_date',
    address='new address',
    phone='new phone',
    money_he_owes_us=money_he_owes_us+'new_money'
  WHERE
    customer_id=id AND address='old address' AND phone='old phone';
@end example

As you can see, this is very efficient and works even if another client
has changed the values in the @code{pay_back} or @code{money_he_owes_us}
columns.

@item
@findex mysql_insert_id()
@findex LAST_INSERT_ID()
In many cases, users have wanted @code{ROLLBACK} and/or @code{LOCK
TABLES} for the purpose of managing unique identifiers for some tables.
This can be handled much more efficiently by using an
@code{AUTO_INCREMENT} column and either the SQL function
@code{LAST_INSERT_ID()} or the C API function @code{mysql_insert_id()}.
@xref{mysql_insert_id, , @code{mysql_insert_id()}}.

@cindex rows, locking
@cindex locking, row-level
You can generally code around row-level locking. Some situations really
need it, but they are very few. @code{InnoDB} tables support row-level
locking. With MyISAM, you can use a flag column in the table and do
something like the following:

@example
UPDATE tbl_name SET row_flag=1 WHERE id=ID;
@end example

MySQL returns 1 for the number of affected rows if the row was
found and @code{row_flag} wasn't already 1 in the original row.

You can think of it as though MySQL Server changed the preceding query to:

@example
UPDATE tbl_name SET row_flag=1 WHERE id=ID AND row_flag <> 1;
@end example
@end itemize


@node ANSI diff Triggers, ANSI diff Foreign Keys, ANSI diff Transactions, Differences from ANSI
@subsubsection Stored Procedures and Triggers

@cindex stored procedures and triggers, defined
@cindex procedures, stored
@cindex triggers, stored

A stored procedure is a set of SQL commands that can be compiled and stored
in the server. Once this has been done, clients don't need to keep re-issuing
the entire query but can refer to the stored procedure. This provides better
performance because the query has to be parsed only once, and less information
needs to be sent between the server and the client. You can also raise the
conceptual level by having libraries of functions in the server.

A trigger is a stored procedure that is invoked when a particular event
occurs.  For example, you can install a stored procedure that is triggered
each time a record is deleted from a transaction table and that automatically
deletes the corresponding customer from a customer table when all his
transactions are deleted.

The planned update language will be able to handle stored procedures.
Our aim is to have stored procedures implemented in MySQL Server around
version 4.1. We are also looking at triggers.


@node ANSI diff Foreign Keys, ANSI diff Views, ANSI diff Triggers, Differences from ANSI
@subsubsection Foreign Keys

@cindex foreign keys
@cindex keys, foreign

Note that foreign keys in SQL are not used to join tables, but are used
mostly for checking referential integrity (foreign key constraints).  If
you want to get results from multiple tables from a @code{SELECT}
statement, you do this by joining tables:

@example
SELECT * FROM table1,table2 WHERE table1.id = table2.id;
@end example

@xref{JOIN, , @code{JOIN}}. @xref{example-Foreign keys}.

In MySQL Server 3.23.44 and up, @code{InnoDB} tables support checking of
foreign key constraints. @xref{InnoDB}. For other table types, MySQL Server
does parse the @code{FOREIGN KEY} syntax in @code{CREATE TABLE}
commands, but without further action being taken.

The @code{FOREIGN KEY} syntax without @code{ON DELETE ...} is mostly
used for documentation purposes. Some ODBC applications may use this
to produce automatic @code{WHERE} clauses, but this is usually easy to
override. @code{FOREIGN KEY} is sometimes used as a constraint check,
but this check is unnecessary in practice if rows are inserted into the
tables in the right order.

In MySQL Server, you can work around the problem of @code{ON DELETE ...} not
being implemented by adding the appropriate @code{DELETE} statement to
an application when you delete records from a table that has a foreign
key. In practice this is as quick (in some cases quicker) and much more
portable than using foreign keys.

In MySQL Server 4.0 you can use multi-table delete to delete rows from many
tables with one command. @xref{DELETE}.

In the near future we will extend the @code{FOREIGN KEY} implementation
so that the information will be saved in the table specification file
and may be retrieved by @code{mysqldump} and ODBC. At a later stage we
will implement the foreign key constraints for applications that can't
easily be coded to avoid them.

Do keep in mind that foreign keys are often misused, which can cause
severe problems. Even when used properly, it is not a magic solution for
the referential integrity problem, although it does make things easier
in some cases.

Some advantages of foreign key enforcement:

@itemize @bullet
@item
Assuming proper design of the relations, foreign key constraints will
make it more difficult for a programmer to introduce an inconsistency
into the database.

@item
Using cascading updates and deletes can simplify the client code.

@item
Properly designed foreign key rules aid in documenting relations
between tables.
@end itemize

Disadvantages:

@itemize @bullet
@item
Mistakes, which are easy to make in designing key relations, can cause
severe problems@-for example, circular rules, or the wrong combination
of cascading deletes.

@item
A properly written application will make sure internally that it is
not violating referential integrity constraints before proceding with
a query. Thus, additional checks on the database level will only slow
down performance for such an application.

@item
It is not uncommon for a DBA to make such a complex topology of
relations that it becomes very difficult, and in some cases impossible,
to back up or restore individual tables.
@end itemize


@node ANSI diff Views, ANSI diff comments, ANSI diff Foreign Keys, Differences from ANSI
@subsubsection Views

@cindex views

It is planned to implement views in MySQL Server around version 4.1.

Views are mostly useful for letting users access a set of relations as one
table (in read-only mode).  Many SQL databases don't allow one to update
any rows in a view, but you have to do the updates in the separate tables.

As MySQL Server is mostly used in applications and on web systems where
the application writer has full control on the database usage, most of
our users haven't regarded views to be very important.
(At least no one has been interested enough in this to be prepared to
finance the implementation of views.)

One doesn't need views in MySQL Server to restrict access to columns,
as MySQL Server has a very sophisticated privilege system.
@xref{Privilege system}.


@node ANSI diff comments,  , ANSI diff Views, Differences from ANSI
@subsubsection @samp{--} as the Start of a Comment

@cindex comments, starting
@cindex starting, comments

Some other SQL databases use @samp{--} to start comments.
MySQL Server has @samp{#} as the start comment character. You can also use
the C comment style @code{/* this is a comment */} with MySQL Server.
@xref{Comments}.

MySQL Server Version 3.23.3 and above support the @samp{--} comment style,
provided the comment is followed by a space.  This is because this
comment style has caused many problems with automatically generated
SQL queries that have used something like the following code, where
we automatically insert the value of the payment for
@code{!payment!}:

@example
UPDATE tbl_name SET credit=credit-!payment!
@end example

Think about what happens if the value of @code{payment} is negative.
Because @code{1--1} is legal in SQL, the consequences of allowing
comments to start with @samp{--} are terrible.

Using our implementation of this method of commenting in MySQL Server
Version 3.23.3 and up, @code{1-- This is a comment} is actually safe.

Another safe feature is that the @code{mysql} command-line client
removes all lines that start with @samp{--}.


The following information is relevant only if you are running a
MySQL version earlier than 3.23.3:

If you have a SQL program in a text file that contains @samp{--}
comments you should use:

@example
shell> replace " --" " #" < text-file-with-funny-comments.sql \
         | mysql database
@end example

instead of the usual:

@example
shell> mysql database < text-file-with-funny-comments.sql
@end example

You can also edit the command file ``in place'' to change the @samp{--}
comments to @samp{#} comments:

@example
shell> replace " --" " #" -- text-file-with-funny-comments.sql
@end example

Change them back with this command:

@example
shell> replace " #" " --" -- text-file-with-funny-comments.sql
@end example


@node Bugs,  , Differences from ANSI, Compatibility
@subsection Known Errors and Design Deficiencies in MySQL

@cindex bugs, known
@cindex errors, known
@cindex design, issues
@cindex known errors

The following problems are known and have a very high priority to get
fixed:

@itemize @bullet
@item
@code{ANALYZE TABLE} on a BDB table may in some case make the table
unusable until one has restarted @code{mysqld}.  When this happens you will
see errors like the following in the MySQL error file:

@example
001207 22:07:56  bdb:  log_flush: LSN past current end-of-log
@end example

@item
Don't execute @code{ALTER TABLE} on a @code{BDB} table on which you are
running multi-statement transactions until all those transactions complete.
(The transaction will probably be ignored.)

@item
@code{ANALYZE TABLE}, @code{OPTIMIZE TABLE}, and @code{REPAIR TABLE} may
cause problems on tables for which you are using @code{INSERT DELAYED}.

@item
Doing a @code{LOCK TABLE ...} and @code{FLUSH TABLES ...} doesn't
guarantee that there isn't a half-finished transaction in progress on the
table.

@item
BDB tables are a bit slow to open. If you have many BDB tables in a
database, it will take a long time to use the @code{mysql} client on the
database if you are not using the @code{-A} option or if you are using
@code{rehash}. This is especially notable when you have a big table
cache.

@item
The current replication protocol cannot deal with @code{LOAD DATA INFILE}
and line terminator characters of more than 1 character.
@end itemize

The following problems are known and will be fixed in due time:

@itemize @bullet

@item
When using @code{SET CHARACTER SET}, one can't use translated
characters in database, table, and column names.

@item
If you have a @code{DECIMAL} column with a  number stored in different
formats (+01.00, 1.00, 01.00), @code{GROUP BY} may regard each value
as a different value.

@item
@code{DELETE FROM merge_table} used without a @code{WHERE}
will only clear the mapping for the table, not delete everything in the
mapped tables.

@item
You cannot build the server in another directory when using
MIT-pthreads. Because this requires changes to MIT-pthreads, we are not
likely to fix this. @xref{MIT-pthreads}.

@item
@code{BLOB} values can't ``reliably'' be used in @code{GROUP BY} or
@code{ORDER BY} or @code{DISTINCT}. Only the first @code{max_sort_length}
bytes (default 1024) are used when comparing @code{BLOB}s in these cases.
This can be changed with the @code{-O max_sort_length} option to
@code{mysqld}. A workaround for most cases is to use a substring:
@code{SELECT DISTINCT LEFT(blob,2048) FROM tbl_name}.

@item
Calculation is done with @code{BIGINT} or @code{DOUBLE} (both are
normally 64 bits long). It depends on the function which precision one
gets. The general rule is that bit functions are done with @code{BIGINT}
precision, @code{IF}, and @code{ELT()} with @code{BIGINT} or @code{DOUBLE}
precision and the rest with @code{DOUBLE} precision.  One should try to
avoid using unsigned long long values if they resolve to be bigger than
63 bits (9223372036854775807) for anything else than bit fields!
MySQL Server 4.0 has better @code{BIGINT} handling than 3.23.

@item
All string columns, except @code{BLOB} and @code{TEXT} columns, automatically
have all trailing spaces removed when retrieved. For @code{CHAR} types this
is okay, and may be regarded as a feature according to ANSI SQL92. The bug is
that in MySQL Server, @code{VARCHAR} columns are treated the same way.

@item
You can only have up to 255 @code{ENUM} and @code{SET} columns in one table.

@item
@code{safe_mysqld} redirects all messages from @code{mysqld} to the
@code{mysqld} log.  One problem with this is that if you execute
@code{mysqladmin refresh} to close and reopen the log,
@code{stdout} and @code{stderr} are still redirected to the old log.
If you use @code{--log} extensively, you should edit @code{safe_mysqld} to
log to @file{'hostname'.err} instead of @file{'hostname'.log} so you can
easily reclaim the space for the old log by deleting the old one and
executing @code{mysqladmin refresh}.

@item
In the @code{UPDATE} statement, columns are updated from left to right.  If
you refer to an updated column, you will get the updated value instead of the
original value. For example:

@example
mysql> UPDATE tbl_name SET KEY=KEY+1,KEY=KEY+1;
@end example

This will update @code{KEY} with @code{2} instead of with @code{1}.

@item
You can't use temporary tables more than once in the same query.
For example, the following doesn't work:

@example
mysql> SELECT * FROM temporary_table, temporary_table AS t2;
@end example

@item
@code{RENAME} doesn't work with @code{TEMPORARY} tables or tables used in a
@code{MERGE} table.

@item
The optimiser may handle @code{DISTINCT} differently if you are using
'hidden' columns in a join or not.  In a join, hidden columns are
counted as part of the result (even if they are not shown) while in
normal queries hidden columns don't participate in the @code{DISTINCT}
comparison.  We will probably change this in the future to never compare
the hidden columns when executing @code{DISTINCT}.

An example of this is:

@example
SELECT DISTINCT mp3id FROM band_downloads
       WHERE userid = 9 ORDER BY id DESC;
@end example

and

@example
SELECT DISTINCT band_downloads.mp3id
       FROM band_downloads,band_mp3
       WHERE band_downloads.userid = 9
       AND band_mp3.id = band_downloads.mp3id
       ORDER BY band_downloads.id DESC;
@end example

In the second case you may in MySQL Server 3.23.x get two identical rows
in the result set (because the hidden @code{id} column may differ).

Note that this happens only for queries where you don't have the
ORDER BY columns in the result, something that you are not allowed
to do in ANSI SQL.

@item
Because MySQL Server allows you to work with table types that don't
support transactions, and thus can't @code{rollback} data, some things
behave a little differently in MySQL Server than in other SQL servers.
This is just to ensure that MySQL Server never needs to do a rollback
for a SQL command.  This may be a little awkward at times as column
values must be checked in the application, but this will actually give
you a nice speed increase as it allows MySQL Server to do some
optimisations that otherwise would be very hard to do.

If you set a column to an incorrect value, MySQL Server will, instead of
doing a rollback, store the @code{best possible value} in the column:

@itemize @minus
@item
If you try to store a value outside the range in a numerical column,
MySQL Server will instead store the smallest or biggest possible value in
the column.

@item
If you try to store a string that doesn't start with a number into a
numerical column, MySQL Server will store 0 into it.

@item
If you try to store @code{NULL} into a column that doesn't take
@code{NULL} values, MySQL Server will store 0 or @code{''} (empty
string) in it instead. (This behavior can, however, be changed with the
-DDONT_USE_DEFAULT_FIELDS compile option.)

@item
MySQL allows you to store some wrong date values into
@code{DATE} and @code{DATETIME} columns (like 2000-02-31 or 2000-02-00).
If the date is totally wrong, MySQL Server will store the special
0000-00-00 date value in the column.

@item
If you set an @code{ENUM} column to an unsupported value, it will be set to
the error value @code{empty string}, with numeric value 0.

@item
If you set a @code{SET} column to an unsupported value, the value will
be ignored.

@end itemize

@item
If you execute a @code{PROCEDURE} on a query that returns an empty set,
in some cases the @code{PROCEDURE} will not transform the columns.

@item
Creation of a table of type @code{MERGE} doesn't check if the underlying
tables are of compatible types.

@item
MySQL Server can't yet handle @code{NaN}, @code{-Inf}, and @code{Inf}
values in double. Using these will cause problems when trying to export
and import data. We should as an intermediate solution change @code{NaN} to
@code{NULL} (if possible) and @code{-Inf} and @code{Inf} to the
minimum respective maximum possible @code{double} value.

@item
@code{LIMIT} on negative numbers are treated as big positive numbers.

@item
If you use @code{ALTER TABLE} to first add a @code{UNIQUE} index to a
table used in a @code{MERGE} table and then use @code{ALTER TABLE} to
add a normal index on the @code{MERGE} table, the key order will be
different for the tables if there was an old key that was not unique in the
table. This is because @code{ALTER TABLE} puts @code{UNIQUE} keys before
normal keys to be able to detect duplicate keys as early as possible.
@end itemize

The following are known bugs in earlier versions of MySQL:

@itemize @bullet
@item
You can get a hung thread if you do a @code{DROP TABLE} on a table that is
one among many tables that is locked with @code{LOCK TABLES}.

@item
In the following case you can get a core dump:

@itemize @minus
@item
Delayed insert handler has pending inserts to a table.

@item
@code{LOCK table} with @code{WRITE}.

@item
@code{FLUSH TABLES}.
@end itemize

@item
Before MySQL Server Version 3.23.2 an @code{UPDATE} that updated a key with
a @code{WHERE} on the same key may have failed because the key was used to
search for records and the same row may have been found multiple times:

@example
UPDATE tbl_name SET KEY=KEY+1 WHERE KEY > 100;
@end example

A workaround is to use:

@example
mysql> UPDATE tbl_name SET KEY=KEY+1 WHERE KEY+0 > 100;
@end example

This will work because MySQL Server will not use an index on expressions in
the @code{WHERE} clause.

@item
Before MySQL Server Version 3.23, all numeric types where treated as
fixed-point fields. That means you had to specify how many decimals
a floating-point field shall have. All results were returned with the
correct number of decimals.
@end itemize

For platform-specific bugs, see the sections about compiling and porting.


@node TODO, Comparisons, Compatibility, Introduction
@section MySQL and The Future (The TODO)

@cindex ToDo list for MySQL

@menu
* TODO MySQL 4.0::              Things That Should be in 4.0
* TODO MySQL 4.1::              Things That Should be in 4.1
* TODO future::                 Things That Must be Done in the Near Future
* TODO sometime::               Things That Have to be Done Sometime
* TODO unplanned::              Things We Have No Plans to do
@end menu

This appendix lists the features that we plan to implement in MySQL Server.

Everything in this list is approximately in the order it will be done.  If you
want to affect the priority order, please register a license or support us and
tell us what you want to have done more quickly. @xref{Licensing and Support}.

The plan is that we in the future will support the full ANSI SQL99
standard, but with a lot of useful extensions. The challenge is to do
this without sacrifying the speed or compromising the code.


@node TODO MySQL 4.0, TODO MySQL 4.1, TODO, TODO
@subsection Things That Should be in 4.0

We are now in the final stages of the development of the MySQL Server
4.0.  server. The target is to quickly implement the rest of the
following features and then shift development to MySQL Server
4.1. @xref{MySQL 4.0 In A Nutshell}.

The news section for 4.0 includes a list of the features we have already
implemented in the 4.0 tree. @xref{News-4.0.x}.

This section lists features not yet implemented in the current version
of MySQL Server 4.0, which will, however, be implemented in later versions
of MySQL 4.0. This being very volatile information, please consider this
list valid only if you are reading it from the MySQL web site
(@uref{http://www.mysql.com/}).

@itemize @bullet
@item
Allow users to change startup options without taking down the server.
@item
Better command-line argument handling.
@item
New key cache, which will give better performance when using many threads.
@item
New table definition file format (@file{.frm} files). This will enable us
to not run out of bits when adding more table options.  One will still
be able to use the old @file{.frm} file format with 4.0.  All newly created
tables will, however, use the new format.

The new file format will enable us to add new column types, more options
for keys, and possibly to store and retrieve @code{FOREIGN KEY} definitions.
@item
@code{SHOW COLUMNS FROM table_name} (used by @code{mysql} client to allow
expansions of column names) should not open the table, only the
definition file. This will require less memory and be much faster.
@item
@code{SET SQL_DEFAULT_TABLE_TYPE=[MyISAM | INNODB | BDB | HEAP]}.
@end itemize


@node TODO MySQL 4.1, TODO future, TODO MySQL 4.0, TODO
@subsection Things That Should be in 4.1

The following features are planned for inclusion into MySQL 4.1.
Note that because we have many developers that are working on different
projects, there will also be many additional features. There is also a
small chance that some of these features will be added to MySQL 4.0.
Some of the work on MySQL 4.1 is already in progress.

@itemize @bullet
@item
Subqueries.
@code{SELECT id FROM t WHERE grp IN (SELECT grp FROM g WHERE u > 100)}
@item
Foreign keys, including cascading delete.
@item
Fail-safe replication.
@item
Replication should work with @code{RAND()} and user variables @code{@@var}.
@item
Online backup with very low performance penalty.  The online backup will
make it easy to add a new replication slave without taking down the
master.
@item
Derived tables:
@example
SELECT a.col1, b.col2
       FROM (SELECT MAX(col1) AS col1 FROM root_table) a,
       other_table b
       WHERE a.col1=b.col1;
@end example

This could be done by automatically creating temporary tables for the
derived tables for the duration of the query.
@item
Allow @code{DELETE} on @code{MyISAM} tables to use the record cache.
To do this, we need to update the threads record cache when we update
the @file{.MYD} file.
@item
When using @code{SET CHARACTER SET} we should translate the whole query
at once and not only strings. This will enable users to use the translated
characters in database, table, and column names.
@item
Add @code{record_in_range()} method to @code{MERGE} tables to be
able to choose the right index when there are many to choose from. We should
also extend the info interface to get the key distribution for each index,
if @code{analyze} is run on all subtables.
@item
@code{RENAME TABLE} on a table used in an active @code{MERGE} table may
corrupt the table.
@item
A faster, smaller embedded MySQL library. (Compatible with the old one.)
@item
Stable openssl support.  (MySQL 4.0 supports rudimentary, not 100% tested,
support for openssl).
@item
Add support for sorting on @code{UNICODE}.
@item
Character set casts and syntax for handling multiple character sets.
@item
Help for all commands from the client.
@item
New faster client/server protocol which will support prepared statements,
bound parameters, and bound result columns, binary transfer of data,
warnings...
@item
Add database and real table name (in case of alias) to the MYSQL_FIELD
structure.
@item
Add options to the client/server protocol to get progress notes
for long running commands.
@item
Implement @code{RENAME DATABASE}. To make this safe for all table handlers,
it should work as follows:
@itemize @bullet
@item
Create the new database.
@item
For every table do a rename of the table to another database, as
we do with the @code{RENAME} command.
@item
Drop the old database.
@end itemize
@item
Add true @code{VARCHAR} support (there is already support for this in
@code{MyISAM}).
@item
Optimise @code{BIT} type to take 1 bit (now @code{BIT} takes 1 char).
@item
New internal file interface change.  This will make all file handling much
more general and make it easier to add extensions like RAID.
(the current implementation is a hack.)
@item
Better in-memory (@code{HEAP}) tables:
@itemize @bullet
@item
Support for B-tree indexes
@item
Dynamic size rows
@item
Faster row handling (less copying)
@end itemize
@end itemize


@node TODO future, TODO sometime, TODO MySQL 4.1, TODO
@subsection Things That Must be Done in the Real Near Future

@itemize @bullet
@item
Atomic multi-table updates@-e.g., @code{update items,month set
items.price=month.price where items.id=month.id;};
@item
Don't allow more than a defined number of threads to run MyISAM recover
at the same time.
@item
Change @code{INSERT ... SELECT} to optionally use concurrent inserts.
@item
Return the original field types() when doing @code{SELECT MIN(column)
... GROUP BY}.
@item
Multiple result sets.
@item
Make it possible to specify @code{long_query_time} with a granularity
in microseconds.
@item
Add a configurable prompt to the @code{mysql} command-line client, with
options like database in use, time and date...
@item
Link the @code{myisampack} code into the server.
@item
Port of the MySQL code to QNX.
@item
Port of the MySQL code to BeOS.
@item
Port of the MySQL clients to LynxOS.
@item
Add a temporary key buffer cache during @code{INSERT/DELETE/UPDATE} so that we
can gracefully recover if the index file gets full.
@item
If you perform an @code{ALTER TABLE} on a table that is symlinked to another
disk, create temporary tables on this disk.
@item
Implement a @code{DATE/DATETIME} type that handles time zone information
properly so that dealing with dates in different time zones is easier.
@item
FreeBSD and MIT-pthreads; do sleeping threads take CPU time?
@item
Check if locked threads take any CPU time.
@item
Fix configure so that one can compile all libraries (like @code{MyISAM})
without threads.
@item
Add an option to periodically flush key pages for tables with delayed keys
if they haven't been used in a while.
@item
Allow join on key parts (optimisation issue).
@item
@code{INSERT SQL_CONCURRENT} and @code{mysqld --concurrent-insert} to do
a concurrent insert at the end of the file if the file is read-locked.
@item
Server-side cursors.
@item
Check if @code{lockd} works with modern Linux kernels; if not, we have
to fix @code{lockd}!  To test this, start @code{mysqld} with
@code{--enable-locking} and run the different fork* test suits. They shouldn't
give any errors if @code{lockd} works.
@item
Allow SQL variables in @code{LIMIT}, like in @code{LIMIT @@a,@@b}.
@item
Allow update of variables in @code{UPDATE} statements. For example:
@code{UPDATE TABLE foo SET @@a=a+b,a=@@a, b=@@a+c}.
@item
Change when user variables are updated so that one can use them with
@code{GROUP BY}, as in the following example:
@code{SELECT id, @@a:=COUNT(*), SUM(sum_col)/@@a FROM table_name GROUP BY id}.
@item
Don't add automatic @code{DEFAULT} values to columns.  Give an error when using
an @code{INSERT} that doesn't contain a column that doesn't have a
@code{DEFAULT}.
@item
Fix @file{libmysql.c} to allow two @code{mysql_query()} commands in a row
without reading results or give a nice error message when one does this.
@item
Check why MIT-pthreads @code{ctime()} doesn't work on some FreeBSD systems.
@item
Add an @code{IMAGE} option to @code{LOAD DATA INFILE} to not update
@code{TIMESTAMP} and @code{AUTO_INCREMENT} fields.
@item
Added @code{LOAD DATE INFILE ... UPDATE} syntax.
@itemize @bullet
@item
For tables with primary keys, if the data contains the primary key,
entries matching that primary key are updated from the remainder of the
columns. However, columns @strong{missing} from the incoming data feed are not
touched.
@item
For tables with primary keys that are missing some part of the key
in the incoming data stream, or that have no primary key, the feed is
treated as a @code{LOAD DATA INFILE ... REPLACE INTO} now.
@end itemize
@item
Make @code{LOAD DATA INFILE} understand syntax like:
@example
LOAD DATA INFILE 'file_name.txt' INTO TABLE tbl_name
     TEXT_FIELDS (text_field1, text_field2, text_field3)
     SET table_field1=CONCAT(text_field1, text_field2),
         table_field3=23
     IGNORE text_field3
@end example
This can be used to skip over extra columns in the text file,
or update columns based on expressions of the read data.
@item
@code{LOAD DATA INFILE 'file_name' INTO TABLE 'table_name' ERRORS TO err_table_name}.
This would cause any errors and warnings to be logged into the
@code{err_table_name} table. That table would have a structure like:

@example
line_number    - line number in datafile
error_message  - the error/warning message
and maybe
data_line      - the line from the datafile
@end example
@item
Automatic output from @code{mysql} to Netscape.
@item
@code{LOCK DATABASES} (with various options.)
@item
@code{DECIMAL} and @code{NUMERIC} types can't read exponential numbers;
@code{Field_decimal::store(const char *from,uint len)} must be recoded
to fix this.
@item
Functions:
ADD_TO_SET(value,set) and REMOVE_FROM_SET(value,set).
@item
Add use of @code{t1 JOIN t2 ON ...} and @code{t1 JOIN t2 USING ...}
Currently, you can only use this syntax with @code{LEFT JOIN}.
@item
Many more variables for @code{show status}.  Records reads and
updates.  Selects on 1 table and selects with joins.  Mean number of
tables in select. Number of @code{ORDER BY} and @code{GROUP BY} queries.
@item
If you abort @code{mysql} in the middle of a query, you should open
another connection and kill the old running query.
Alternatively, an attempt should be made to detect this in the server.
@item
Add a handler interface for table information so that you can use it as a system
table. This would be a bit slow if you requested information about all tables,
but very flexible.  @code{SHOW INFO FROM tbl_name} for basic table information
should be implemented.
@item
@code{NATURAL JOIN}.
@item
Allow @code{SELECT a FROM crash_me LEFT JOIN crash_me2 USING (a)}; in this
case @code{a} is assumed to come from the @code{crash_me} table.
@item
Fix so that @code{ON} and @code{USING} works with the @code{JOIN}
join type.
@item
Oracle-like @code{CONNECT BY PRIOR ...} to search hierarchy structures.
@item
@code{mysqladmin copy database new-database}; requires @code{COPY}
command to be added to @code{mysqld}.
@item
Processlist should show number of queries/threads.
@item
@code{SHOW HOSTS} for printing information about the hostname cache.
@item
@code{DELETE} and @code{REPLACE} options to the @code{UPDATE} statement
(this will delete rows when one gets a duplicate key error while updating).
@item
Change the format of @code{DATETIME} to store fractions of seconds.
@item
Add all missing ANSI92 and ODBC 3.0 types.
@item
Change table names from empty strings to @code{NULL} for calculated columns.
@item
Don't use @code{Item_copy_string} on numerical values to avoid
number->string->number conversion in case of:
@code{SELECT COUNT(*)*(id+0) FROM table_name GROUP BY id}
@item
Make it possible to use the new GNU regexp library instead of the current
one (the GNU library should be much faster than the old one).
@item
Change so that @code{ALTER TABLE} doesn't abort clients
that execute @code{INSERT DELAYED}.
@item
Fix so that when columns are referenced in an @code{UPDATE} clause,
they contain the old values from before the update started.
@item
Add simulation of @code{pread()}/@code{pwrite()} on Windows to enable
concurrent inserts.
@item
A logfile analyser that could parse out information about which tables
are hit most often, how often multi-table joins are executed, etc.  It
should help users identify areas or table design that could be optimised
to execute much more efficient queries.
@item
Add @code{SUM(DISTINCT)}.
@item
Add @code{ANY()}, @code{EVERY()}, and @code{SOME()} group functions. In
ANSI SQL these work only on boolean columns, but we can extend these to
work on any columns/expressions by applying: value == 0 -> FALSE and
value <> 0 -> TRUE.
@item
Fix that the type for @code{MAX(column)} is the same as the column type:
@example
mysql> CREATE TABLE t1 (a DATE);
mysql> INSERT INTO t1 VALUES (NOW());
mysql> CREATE TABLE t2 SELECT MAX(a) FROM t1;
mysql> SHOW COLUMNS FROM t2;
@end example
@item
Come up with a nice syntax for a statement that will @code{UPDATE} the row
if it exists and @code{INSERT} a new row if the row didn't exist
(like @code{REPLACE} works with @code{INSERT} / @code{DELETE}).
@end itemize


@node TODO sometime, TODO unplanned, TODO future, TODO
@subsection Things That Have to be Done Sometime

@itemize @bullet
@item
Implement function: @code{get_changed_tables(timeout,table1,table2,...)}.
@item
Change reading through tables to use memmap when possible. Now only
compressed tables use memmap.
@item
Add a new privilege @code{Show_priv} for @code{SHOW} commands.
@item
Make the automatic timestamp code nicer.  Add timestamps to the update
log with @code{SET TIMESTAMP=#;}.
@item
Use read/write mutex in some places to get more speed.
@item
Full foreign key support. One probably wants to implement a procedural
language first.
@item
Simple views (first on one table, later on any expression).
@item
Automatically close some tables if a table, temporary table, or temporary files
gets error 23 (not enough open files).
@item
When one finds a field=#, change all occurrences of field to #.  Now this
is only done for some simple cases.
@item
Change all const expressions with calculated expressions if possible.
@item
Optimise key = expression. At the moment only key = field or key =
constant are optimised.
@item
Join some of the copy functions for nicer code.
@item
Change @file{sql_yacc.yy} to an inline parser to reduce its size and get
better error messages (5 days).
@item
Change the parser to use only one rule per different number of arguments
in function.
@item
Use of full calculation names in the order part (for ACCESS97).
@item
@code{MINUS}, @code{INTERSECT}, and @code{FULL OUTER JOIN}.
(Currently @code{UNION} [in 4.0] and @code{LEFT OUTER JOIN} are supported.)
@item
@code{SQL_OPTION MAX_SELECT_TIME=#} to put a time limit on a query.
@item
Make the update log to a database.
@item
Negative @code{LIMIT} to retrieve data from the end.
@item
Alarm around client connect/read/write functions.
@item
Please note the changes to @code{safe_mysqld}: according to FSSTND (which
Debian tries to follow) PID files should go into @file{/var/run/<progname>.pid}
and log files into @file{/var/log}. It would be nice if you could put the
"DATADIR" in the first declaration of "pidfile" and "log", so the
placement of these files can be changed with a single statement.
@item
Allow a client to request logging.
@item
Add use of @code{zlib()} for @code{gzip}-ed files to @code{LOAD DATA INFILE}.
@item
Fix sorting and grouping of @code{BLOB} columns (partly solved now).
@item
Stored procedures. Triggers are also being looked at.
@item
A simple (atomic) update language that
can be used to write loops and such in the MySQL server.
@item
Change to use semaphores when counting threads.  One should first implement
a semaphore library to MIT-pthreads.
@item
Don't assign a new @code{AUTO_INCREMENT} value when one sets a column to 0.
Use @code{NULL} instead.
@item
Add full support for @code{JOIN} with parentheses.
@item
As an alternative for one thread/connection manage a pool of threads
to handle the queries.
@item
Allow one to get more than one lock with @code{GET_LOCK}.  When doing this,
one must also handle the possible deadlocks this change will introduce.
@end itemize

Time is given according to amount of work, not real time.


@node TODO unplanned,  , TODO sometime, TODO
@subsection Things We Have No Plans to do

@itemize @bullet
@item
Nothing; we aim toward full ANSI 92/ANSI 99 compliancy.
@end itemize


@node Comparisons,  , TODO, Introduction
@section How MySQL Compares to Other Databases

@cindex databases, MySQL vs. others
@cindex comparisons, MySQL vs. others

Our users have successfully run their own benchmarks against a number
of @code{Open Source} and traditional database servers.
We are aware of tests against @code{Oracle} server, @code{DB/2} server,
@code{Microsoft SQL Server}, and other commercial products.
Due to legal reasons we are restricted from publishing some of those
benchmarks in our reference manual.

This section includes a comparison with @code{mSQL} for historical
reasons and with @code{PostgreSQL} as it is also an @code{Open Source}
database. If you have benchmark results that we can publish, please
contact us at @email{benchmarks@@mysql.com}.

For comparative lists of all supported functions and types as well
as measured operational limits of many different database systems,
see the @code{crash-me} web page at
@uref{http://www.mysql.com/information/crash-me.php}.

@menu
* Compare mSQL::                How MySQL compares to @code{mSQL}
* Compare PostgreSQL::          How MySQL Compares to @code{PostgreSQL}
@end menu


@node Compare mSQL, Compare PostgreSQL, Comparisons, Comparisons
@subsection How MySQL Compares to @code{mSQL}

@cindex mSQL, MySQL vs mSQL, overview
@table @strong
@item Performance

For a true comparison of speed, consult the growing MySQL benchmark
suite. @xref{MySQL Benchmarks}.

Because there is no thread creation overhead, a small parser, few
features, and simple security, @code{mSQL} should be quicker at:

@itemize @bullet
@item
Tests that perform repeated connects and disconnects, running a very
simple query during each connection.

@item
@code{INSERT} operations into very simple tables with few columns and keys.

@item
@code{CREATE TABLE} and @code{DROP TABLE}.

@item
@code{SELECT} on something that isn't an index. (A table scan is very
easy.)
@end itemize

Because these operations are so simple, it is hard to be better at
them when you have a higher startup overhead. After the connection
is established, MySQL Server should perform much better.

On the other hand, MySQL Server is much faster than @code{mSQL} (and
most other SQL implementations) on the following:

@itemize @bullet
@item
Complex @code{SELECT} operations.

@item
Retrieving large results (MySQL Server has a better, faster, and safer
protocol).

@item
Tables with variable-length strings because MySQL Server has more efficient
handling and can have indexes on @code{VARCHAR} columns.

@item
Handling tables with many columns.

@item
Handling tables with large record lengths.

@item
@code{SELECT} with many expressions.

@item
@code{SELECT} on large tables.

@item
Handling many connections at the same time. MySQL Server is fully
multi-threaded. Each connection has its own thread, which means that
no thread has to wait for another (unless a thread is modifying
a table another thread wants to access). In @code{mSQL}, once one
connection is established, all others must wait until the first has
finished, regardless of whether the connection is running a query
that is short or long. When the first connection terminates, the
next can be served, while all the others wait again, etc.

@item
Joins.
@code{mSQL} can become pathologically slow if you change the order of
tables in a @code{SELECT}. In the benchmark suite, a time more than
15,000 times slower than MySQL Server was seen.  This is due to @code{mSQL}'s
lack of a join optimiser to order tables in the optimal order.
However, if you put the tables in exactly the right order in
@code{mSQL}2 and the @code{WHERE} is simple and uses index columns,
the join will be relatively fast!
@xref{MySQL Benchmarks}.

@item
@code{ORDER BY} and @code{GROUP BY}.

@item
@code{DISTINCT}.

@item
Using @code{TEXT} or @code{BLOB} columns.
@end itemize

@item SQL Features

@itemize @bullet
@item @code{GROUP BY} and @code{HAVING}.
@code{mSQL} does not support @code{GROUP BY} at all.
MySQL Server supports a full @code{GROUP BY} with both @code{HAVING} and
the following functions: @code{COUNT()}, @code{AVG()}, @code{MIN()},
@code{MAX()}, @code{SUM()}, and @code{STD()}.  @code{COUNT(*)} is
optimised to return very quickly if the @code{SELECT} retrieves from
one table, no other columns are retrieved, and there is no
@code{WHERE} clause.  @code{MIN()} and @code{MAX()} may take string
arguments.

@item @code{INSERT} and @code{UPDATE} with calculations.
MySQL Server can do calculations in an @code{INSERT} or @code{UPDATE}.
For example:

@example
mysql> UPDATE SET x=x*10+y WHERE x<20;
@end example

@item Aliasing.
MySQL Server has column aliasing.

@item Qualifying column names.
In MySQL Server, if a column name is unique among the tables used in a
query, you do not have to use the full qualifier.

@item @code{SELECT} with functions.
MySQL Server has many functions (too many to list here; see @ref{Functions}).

@end itemize

@item Disk Space Efficiency
That is, how small can you make your tables?

MySQL Server has very precise types, so you can create tables that take
very little space. An example of a useful MySQL datatype is the
@code{MEDIUMINT} that is 3 bytes long. If you have 100 million
records, saving even 1 byte per record is very important.

@code{mSQL2} has a more limited set of column types, so it is
more difficult to get small tables.

@item Stability
This is harder to judge objectively. For a discussion of MySQL Server
stability, see @ref{Stability}.

We have no experience with @code{mSQL} stability, so we cannot say
anything about that.

@item Price
Another important issue is the license. MySQL Server has a
more flexible license than @code{mSQL}, and is also less expensive
than @code{mSQL}.  Whichever product you choose to use, remember to
at least consider paying for a license or e-mail support.

@item Perl Interfaces
MySQL Server has basically the same interfaces to Perl as @code{mSQL} with
some added features.

@item JDBC (Java)
MySQL Server currently has a lot of different JDBC drivers:

@itemize @bullet
@item
The mm driver: a type 4 JDBC driver by Mark Matthews
@email{mmatthew@@ecn.purdue.edu}.  This is released under the LGPL.

@item
The Resin driver: this is a commercial JDBC driver released under open
source.  @uref{http://www.caucho.com/projects/jdbc-mysql/index.xtp}

@item
The gwe driver: a Java interface by GWE technologies (not supported anymore).

@item
The jms driver: an improved gwe driver by Xiaokun Kelvin ZHU
@email{X.Zhu@@brad.ac.uk} (not supported anymore).

@item
The twz driver: a type 4 JDBC driver by Terrence W. Zellers
@email{zellert@@voicenet.com}.  This is commercial but is free for private
and educational use (not supported anymore).
@end itemize

The recommended driver is the mm driver. The Resin driver may also be
good (at least the benchmarks look good), but we haven't received that
much information about this yet.

We know that @code{mSQL} has a JDBC driver, but we have too little
experience with it to compare.

@item Rate of Development
MySQL Server has a small core team of developers, but we are quite
used to coding C and C++ very rapidly. Because threads, functions,
@code{GROUP BY}, and so on are still not implemented in @code{mSQL}, it
has a lot of catching up to do. To get some perspective on this, you
can view the @code{mSQL} @file{HISTORY} file for the last year and
compare it with the News section of the MySQL Reference Manual
(@pxref{News}). It should be pretty obvious which one has developed
most rapidly.

@item Utility Programs
Both @code{mSQL} and MySQL Server have many interesting third-party
tools. Because it is very easy to port upward (from @code{mSQL} to
MySQL Server), almost all the interesting applications that are available for
@code{mSQL} are also available for MySQL Server.

MySQL Server comes with a simple @code{msql2mysql} program that fixes
differences in spelling between @code{mSQL} and MySQL Server for the
most-used C API functions.
For example, it changes instances of @code{msqlConnect()} to
@code{mysql_connect()}. Converting a client program from @code{mSQL} to
MySQL Server usually requires only minor effort.
@end table

@menu
* Using mSQL tools::            How to convert @code{mSQL} tools for MySQL
* Protocol differences::        How @code{mSQL} and MySQL Client/Server Communications Protocols Differ
* Syntax differences::          How @code{mSQL} 2.0 SQL Syntax Differs from MySQL
@end menu


@node Using mSQL tools, Protocol differences, Compare mSQL, Compare mSQL
@subsubsection How to Convert @code{mSQL} Tools for MySQL

@cindex MySQL tools, conversion
@cindex converting, tools
@cindex tools, converting

According to our experience, it doesn't take long to convert tools
such as @code{msql-tcl} and @code{msqljava} that use the
@code{mSQL} C API so that they work with the MySQL C API.

The conversion procedure is:

@enumerate
@item
Run the shell script @code{msql2mysql} on the source. This requires
the @code{replace} program, which is distributed with MySQL Server.

@item
Compile.

@item
Fix all compiler errors.
@end enumerate

Differences between the @code{mSQL} C API and the MySQL C API are:

@itemize @bullet
@item
MySQL Server uses a @code{MYSQL} structure as a connection type (@code{mSQL}
uses an @code{int}).

@item
@code{mysql_connect()} takes a pointer to a @code{MYSQL} structure as a
parameter.  It is easy to define one globally or to use @code{malloc()}
to get one.  @code{mysql_connect()} also takes two parameters for
specifying the user and password.  You may set these to
@code{NULL, NULL} for default use.

@item
@code{mysql_error()} takes the @code{MYSQL} structure as a parameter.
Just add the parameter to your old @code{msql_error()} code if you are
porting old code.

@item
MySQL Server returns an error number and a text error message for all
errors. @code{mSQL} returns only a text error message.

@item
Some incompatibilities exist as a result of MySQL Server supporting
multiple connections to the server from the same process.
@end itemize


@node Protocol differences, Syntax differences, Using mSQL tools, Compare mSQL
@subsubsection How @code{mSQL} and MySQL Client/Server Communications Protocols Differ

@cindex communications protocols
@cindex mSQL vs. MySQL, protocol

There are enough differences that it is impossible
(or at least not easy) to support both.

The most significant ways in which the MySQL protocol differs
from the @code{mSQL} protocol are listed here:

@itemize @bullet
@item
A message buffer may contain many result rows.

@item
The message buffers are dynamically enlarged if the query or the
result is bigger than the current buffer, up to a configurable server
and client limit.

@item
All packets are numbered to catch duplicated or missing packets.

@item
All column values are sent in ASCII. The lengths of columns and rows
are sent in packed binary coding (1, 2, or 3 bytes).

@item
MySQL can read in the result unbuffered (without having to store the
full set in the client).

@item
If a single read/write takes more than 30 seconds, the server closes
the connection.

@item
If a connection is idle for 8 hours, the server closes the connection.
@end itemize


@node Syntax differences,  , Protocol differences, Compare mSQL
@subsubsection How @code{mSQL} 2.0 SQL Syntax Differs from MySQL

@noindent
@strong{Column types}

@table @code
@item MySQL Server
Has the following additional types (among others;
@pxref{CREATE TABLE, , @code{CREATE TABLE}}):
@itemize @bullet
@item
@c FIX bad lingo, needs rephrasing
@code{ENUM} type for one of a set of strings.
@item
@c FIX bad lingo, needs rephrasing
@code{SET} type for many of a set of strings.
@item
@code{BIGINT} type for 64-bit integers.
@end itemize
@item
MySQL Server also supports
the following additional type attributes:
@itemize @bullet
@item
@code{UNSIGNED} option for integer and floating-point columns.
@item
@code{ZEROFILL} option for integer columns.
@item
@code{AUTO_INCREMENT} option for integer columns that are a
@code{PRIMARY KEY}.
@xref{mysql_insert_id, , @code{mysql_insert_id()}}.
@item
@code{DEFAULT} value for all columns.
@end itemize
@item mSQL2
@code{mSQL} column types correspond to the MySQL types shown in the following table:
@multitable @columnfractions .15 .70
@item @code{mSQL} @strong{type} @tab @strong{Corresponding MySQL type}
@item @code{CHAR(len)} @tab @code{CHAR(len)}
@item @code{TEXT(len)} @tab @code{TEXT(len)}. @code{len} is the maximal length.
And @code{LIKE} works.
@item @code{INT} @tab @code{INT}. With many more options!
@item @code{REAL} @tab @code{REAL}. Or @code{FLOAT}. Both 4- and 8-byte versions are available.
@item @code{UINT} @tab @code{INT UNSIGNED}
@item @code{DATE} @tab @code{DATE}. Uses ANSI SQL format rather than @code{mSQL}'s own format.
@item @code{TIME} @tab @code{TIME}
@item @code{MONEY} @tab @code{DECIMAL(12,2)}. A fixed-point value with two decimals.
@end multitable
@end table

@noindent
@strong{Index Creation}

@table @code
@item MySQL Server
Indexes may be specified at table creation time with the @code{CREATE TABLE}
statement.
@item mSQL
Indexes must be created after the table has been created, with separate
@code{CREATE INDEX} statements.
@end table

@noindent
@strong{To Insert a Unique Identifier into a Table}

@table @code
@item MySQL Server
Use @code{AUTO_INCREMENT} as a column type
specifier.
@xref{mysql_insert_id, , @code{mysql_insert_id()}}.
@item mSQL
Create a @code{SEQUENCE} on a table and select the @code{_seq} column.
@end table

@noindent
@strong{To Obtain a Unique Identifier for a Row}

@table @code
@item MySQL Server
Add a @code{PRIMARY KEY} or @code{UNIQUE} key to the table and use this.
New in Version 3.23.11: If the @code{PRIMARY} or @code{UNIQUE} key consists of only one
column and this is of type integer, one can also refer to it as
@code{_rowid}.
@item mSQL
Use the @code{_rowid} column. Observe that @code{_rowid} may change over time
depending on many factors.
@end table

@noindent
@strong{To Get the Time a Column Was Last Modified}

@table @code
@item MySQL Server
Add a @code{TIMESTAMP} column to the table. This column is automatically set
to the current date and time for @code{INSERT} or @code{UPDATE} statements if
you don't give the column a value or if you give it a @code{NULL} value.

@item mSQL
Use the @code{_timestamp} column.
@end table

@noindent
@strong{@code{NULL} Value Comparisons}

@table @code
@item MySQL Server
MySQL Server follows
ANSI SQL, and a comparison with @code{NULL} is always @code{NULL}.
@item mSQL
In @code{mSQL}, @code{NULL = NULL} is TRUE.  You
must change @code{=NULL} to @code{IS NULL} and @code{<>NULL} to
@code{IS NOT NULL} when porting old code from @code{mSQL} to MySQL Server.
@end table

@noindent
@strong{String Comparisons}

@table @code
@item MySQL Server
Normally, string comparisons are performed in case-independent fashion with
the sort order determined by the current character set (ISO-8859-1 Latin1 by
default).  If you don't like this, declare your columns with the
@code{BINARY} attribute, which causes comparisons to be done according to the
ASCII order used on the MySQL server host.
@item mSQL
All string comparisons are performed in case-sensitive fashion with
sorting in ASCII order.
@end table

@noindent
@strong{Case-insensitive Searching}

@table @code
@item MySQL Server
@code{LIKE} is a case-insensitive or case-sensitive operator, depending on
the columns involved. If possible, MySQL uses indexes if the
@code{LIKE} argument doesn't start with a wildcard character.
@item mSQL
Use @code{CLIKE}.
@end table

@noindent
@strong{Handling of Trailing Spaces}

@table @code
@item MySQL Server
Strips all spaces at the end of @code{CHAR} and @code{VARCHAR}
columns. Use a @code{TEXT} column if this behavior is not desired.
@item mSQL
Retains trailing space.
@end table

@noindent
@strong{@code{WHERE} Clauses}

@table @code
@item MySQL Server
MySQL correctly prioritises everything (@code{AND} is evaluated
before @code{OR}). To get @code{mSQL} behavior in MySQL Server, use
parentheses (as shown in an example later in this section).
@item mSQL
Evaluates everything from left to right. This means that some logical
calculations with more than three arguments cannot be expressed in any
way. It also means you must change some queries when you upgrade to
MySQL Server. You do this easily by adding parentheses.  Suppose you
have the following @code{mSQL} query:
@example
mysql> SELECT * FROM table WHERE a=1 AND b=2 OR a=3 AND b=4;
@end example
To make MySQL Server evaluate this the way that @code{mSQL} would,
you must add parentheses:
@example
mysql> SELECT * FROM table WHERE (a=1 AND (b=2 OR (a=3 AND (b=4))));
@end example
@end table

@noindent
@strong{Access Control}

@table @code
@item MySQL Server
Has tables to store grant (permission) options per user, host, and
database. @xref{Privileges}.
@item mSQL
Has a file @file{mSQL.acl} in which you can grant read/write privileges for
users.
@end table


@node Compare PostgreSQL,  , Compare mSQL, Comparisons
@subsection How MySQL Compares to @code{PostgreSQL}

@cindex PostgreSQL vs. MySQL, overview

When reading the following, please note that both products are continually
evolving.  We at MySQL AB and the PostgreSQL developers are both working
on making our respective databases as good as possible, so we are both a
serious alternative to any commercial database.

The following comparison is made by us at MySQL AB.  We have tried to be
as accurate and fair as possible, but although we know MySQL Server thoroughly,
we don't have a full knowledge of all PostgreSQL features, so we may have
got some things wrong.  We will, however, correct these when they come to our
attention.

We would first like to note that PostgreSQL and MySQL Server are both widely used
products, but with different design goals, even if we are both striving
toward ANSI SQL compliancy.  This means that for some applications MySQL Server
is more suited, while for others PostgreSQL is more suited.  When choosing
which database to use, you should first check if the database's feature set
satisfies your application.  If you need raw speed, MySQL Server is probably your
best choice.  If you need some of the extra features that only PostgreSQL
can offer, you should use @code{PostgreSQL}.

@menu
* MySQL-PostgreSQL goals::      MySQL and PostgreSQL development strategies
* MySQL-PostgreSQL features::   Featurewise Comparison of MySQL and PostgreSQL
* MySQL-PostgreSQL benchmarks::  Benchmarking MySQL and PostgreSQL
@end menu


@node MySQL-PostgreSQL goals, MySQL-PostgreSQL features, Compare PostgreSQL, Compare PostgreSQL
@subsubsection MySQL and PostgreSQL development strategies

@cindex PostgreSQL vs. MySQL, strategies
When adding things to MySQL Server we take pride to do an optimal, definite
solution.  The code should be so good that we shouldn't have any need to
change it in the foreseeable future.  We also do not like to sacrifice
speed for features but instead will do our utmost to find a solution
that will give maximal throughput.  This means that development will take
a little longer, but the end result will be well worth this.  This kind
of development is only possible because all server code are checked by
one of a few (currently two) persons before it's included in the
MySQL server.

We at MySQL AB believe in frequent releases to be able to push out new
features quickly to our users.  Because of this we do a new small release
about every three weeks, and a major branch every year.  All releases are
thoroughly tested with our testing tools on a lot of different platforms.

PostgreSQL is based on a kernel with lots of contributors. In this setup
it makes sense to prioritise adding a lot of new features, instead of
implementing them optimally, because one can always optimise things
later if there arises a need for this.

Another big difference between MySQL Server and PostgreSQL is that
nearly all of the code in the MySQL server is coded by developers that
are employed by MySQL AB and are still working on the server code. The
exceptions are the transaction engines and the regexp library.

This is in sharp contrast to the PostgreSQL code, the majority of
which is coded by a big group of people with different backgrounds.
It was only recently that the PostgreSQL developers announced that their
current developer group had finally had time to take a look at all
the code in the current PostgreSQL release.

Both of the aforementioned development methods have their own merits and drawbacks.
We here at MySQL AB think, of course, that our model is better because our
model gives better code consistency, more optimal and reusable code, and
in our opinion, fewer bugs.  Because we are the authors of the MySQL server
code, we are better able to coordinate new features and releases.


@node MySQL-PostgreSQL features, MySQL-PostgreSQL benchmarks, MySQL-PostgreSQL goals, Compare PostgreSQL
@subsubsection Featurewise Comparison of MySQL and PostgreSQL

@cindex PostgreSQL vs. MySQL, features

On the crash-me page
(@uref{http://www.mysql.com/information/crash-me.php})
you can find a list of those database constructs and limits that
one can detect automatically with a program.  Note, however, that a lot of
the numerical limits may be changed with startup options for their respective
databases.  This web page is, however, extremely useful when you want to
ensure that your applications work with many different databases or
when you want to convert your application from one database to another.

MySQL Server offers the following advantages over PostgreSQL:

@itemize @bullet
@item
@code{MySQL} Server is generally much faster than PostgreSQL.  MySQL
4.0.1 also has a query cache that can boost up the query speed for
mostly-read-only sites many times.

@item
MySQL has a much larger user base than PostgreSQL. Therefore, the code is
tested more and has historically proven more stable than PostgreSQL.
MySQL Server is used more in production environments than PostgreSQL,
mostly thanks to the fact that MySQL AB, formerly TCX DataKonsult AB, has
provided top-quality commercial support for MySQL Server from the day it
was released, whereas until recently PostgreSQL was unsupported.

@item
MySQL Server works better on Windows than PostgreSQL does.  MySQL Server
runs as a native Windows application (a service on NT/2000/XP),
while PostgreSQL is run under the @code{Cygwin} emulation.  We have
heard that PostgreSQL is not yet that stable on Windows but we haven't
been able to verify this ourselves.

@item
MySQL has more APIs to other languages and is supported by more
existing programs than PostgreSQL.  @xref{Contrib}.

@item
MySQL Server works on 24/7 heavy-duty systems.  In most circumstances
you never have to run any cleanups on MySQL Server.  PostgreSQL doesn't
yet support 24/7 systems because you have to run @code{VACUUM}
once in a while to reclaim space from @code{UPDATE} and @code{DELETE}
commands and to perform statistics analyses that are critical to get
good performance with PostgreSQL.  @code{VACUUM} is also needed after
adding a lot of new rows to a table.  On a busy system with lots of changes,
@code{VACUUM} must be run very frequently, in the worst cases even
many times a day.  During the @code{VACUUM} run, which may take hours
if the database is big, the database is, from a production standpoint,
practically dead. Please note: in PostgreSQL version 7.2, basic vacuuming
no longer locks tables, thus allowing normal user access during the vacuum.
A new @code{VACUUM FULL} command does old-style vacuum by locking the table
and shrinking the on-disk copy of the table.

@item
MySQL replication has been thoroughly tested, and is used by sites like:
@itemize @minus
@item Yahoo Finance (@uref{http://finance.yahoo.com/})
@item Mobile.de (@uref{http://www.mobile.de/})
@item Slashdot (@uref{http://www.slashdot.org/})
@end itemize

@item
Included in the MySQL distribution are two different testing suites,
@file{mysql-test-run} and crash-me
(@uref{http://www.mysql.com/information/crash-me.php}), as well
as a benchmark suite.  The test system is actively updated with code to
test each new feature and almost all reproduceable bugs that have come to
our attention.  We test MySQL Server with these on a lot of platforms before
every release.  These tests are more sophisticated than anything we have
seen from PostgreSQL, and they ensure that the MySQL Server is kept to a high
standard.

@item
There are far more books in print about MySQL Server than about PostgreSQL.
O'Reilly, SAMS, Que, and New Riders are all major publishers with books
about MySQL.  All MySQL features are also documented in the MySQL online
manual because when a new feature is implemented, the MySQL developers
are required to document it before it's included in the source.

@item
MySQL Server supports more of the standard ODBC functions than @code{PostgreSQL}.

@item
MySQL Server has a much more sophisticated @code{ALTER TABLE}.

@item
MySQL Server has support for tables without transactions for applications that
need all the speed they can get. The tables may be memory-based, @code{HEAP}
tables or disk based @code{MyISAM}.  @xref{Table types}.

@item
MySQL Server has support for two different table handlers that support
transactions, @code{InnoDB}, and @code{BerkeleyDB}.  Because every
transaction engine performs differently under different conditions, this
gives the application writer more options to find an optimal solution for
his or her setup, if need be per individual table.  @xref{Table types}.

@item
@code{MERGE} tables gives you a unique way to instantly make a view over
a set of identical tables and use these as one.  This is perfect for
systems where you have log files that you order, for example, by month.
@xref{MERGE}.

@item
The option to compress read-only tables, but still have direct access to
the rows in the table, gives you better performance by minimising disk
reads.  This is very useful when you are archiving things.
@xref{myisampack, , @code{myisampack}}.

@item
MySQL Server has internal support for full-text search. @xref{Fulltext Search}.

@item
You can access many databases from the same connection (depending, of course,
on your privileges).

@item
MySQL Server is coded from the start to be multi-threaded, while
PostgreSQL uses processes.  Context switching and access to common
storage areas is much faster between threads than between separate
processes. This gives MySQL Server a big speed advantage in multi-user
applications and also makes it easier for MySQL Server to take full
advantage of symmetric multiprocessor (SMP) systems.

@item
MySQL Server has a much more sophisticated privilege system than
PostgreSQL.  While PostgreSQL only supports @code{INSERT},
@code{SELECT}, and @code{UPDATE/DELETE} grants per user on a database or
a table, MySQL Server allows you to define a full set of different
privileges on the database, table, and column level.  MySQL Server also
allows you to specify the privilege on host and user combinations.
@xref{GRANT}.

@item
MySQL Server supports a compressed client/server protocol which improves
performance over slow links.

@item
MySQL Server employs a ``table handler'' concept, and is the only relational
database we know of built around this concept.  This allows different
low-level table types to be called from the SQL engine, and each table
type can be optimised for different performance characteristics.

@item
All MySQL table types (except @code{InnoDB}) are implemented as files
(one table per file), which makes it really easy to back up, move, delete,
and even symlink databases and tables, even when the server is down.

@item
Tools to repair and optimise @code{MyISAM} tables (the most common
MySQL table type).  A repair tool is only needed when a physical corruption
of a datafile happens, usually from a hardware failure.  It allows a
majority of the data to be recovered.

@item
Upgrading MySQL Server is painless.  When you are upgrading MySQL
Server, you don't need to dump/restore your data, as you have to do with
most PostgreSQL upgrades.
@end itemize

Drawbacks with MySQL Server compared to PostgreSQL:

@itemize @bullet
@item
The transaction support in MySQL Server is not yet as well tested as
PostgreSQL's system.

@item
Because MySQL Server uses threads, which are not yet flawless on many OSes, one
must either use binaries from @uref{http://www.mysql.com/downloads/}, or
carefully follow our instructions on
@uref{http://www.mysql.com/doc/I/n/Installing_source.html} to get an
optimal binary that works in all cases.

@item
Table locking, as used by the non-transactional @code{MyISAM} tables, is
in many cases faster than page locks, row locks, or versioning.  The
drawback, however, is that if one doesn't take into account how table
locks work, a single long-running query can block a table for updates
for a long time. This can usually be avoided when designing the
application.  If not, one can always switch the trouble table to use one
of the transactional table types. @xref{Table locking}.

@item
With UDF (user-defined functions) one can extend MySQL Server with both normal
SQL functions and aggregates, but this is not yet as easy or as flexible
as in PostgreSQL.  @xref{Adding functions}.

@item
Updates that run over multiple tables are harder to do in MySQL Server.
This will, however, be fixed in MySQL Server 4.0.2 with multi-table
@code{UPDATE} and in MySQL Server 4.1 with subselects.  In MySQL Server
4.0 one can use multi-table deletes to delete from many tables at the
same time. @xref{DELETE}.
@end itemize

PostgreSQL currently offers the following advantages over MySQL Server:

Note that because we know the MySQL road map, we have included in the
following table the version when MySQL Server should support this feature.
Unfortunately we couldn't do this for previous comparisons, because we
don't know the PostgreSQL roadmap.

@multitable @columnfractions .30 .30
@item @strong{Feature}                     @tab @strong{MySQL version}
@item Subselects                           @tab 4.1
@item Foreign keys                         @tab 4.1
@item Views                                @tab 5.0
@item Stored procedures                    @tab 5.0
@item Triggers                             @tab 5.0
@item Unions                               @tab 4.0
@item Full join                            @tab 4.1
@item Constraints                          @tab 4.1 or 5.0
@item Cursors                              @tab 4.1 or 5.0
@item R-trees                              @tab 4.1 (for MyISAM tables)
@item Inherited tables                     @tab Not planned
@item Extensible type system               @tab Not planned
@end multitable

Other reasons someone may consider using PostgreSQL:

@itemize @bullet
@item
Standard usage in PostgreSQL is closer to ANSI SQL in some cases.

@item
One can speed up PostgreSQL by coding things as stored procedures.

@item
For geographical data, R-trees make PostgreSQL better than MySQL Server.
(note: MySQL version 4.1 will have R-trees for MyISAM tables).

@item
The PostgreSQL optimiser can do some optimisation that the current MySQL
optimiser can't do.  Most notable is doing joins when you don't have the
proper keys in place and doing a join where you are using different keys
combined with OR.  The MySQL benchmark suite at
@uref{http://www.mysql.com/information/benchmarks.html} shows you what
kind of constructs you should watch out for when using different
databases.

@item
PostgreSQL has a bigger team of developers that contribute to the server.
@end itemize

Drawbacks with PostgreSQL compared to MySQL Server:

@itemize @bullet
@item
@code{VACUUM} makes PostgreSQL hard to use in a 24/7 environment.

@item
Only transactional tables.

@item
Much slower @code{INSERT}, @code{DELETE}, and @code{UPDATE}.
@end itemize

For a complete list of drawbacks, you should also examine the first table
in this section.


@node MySQL-PostgreSQL benchmarks,  , MySQL-PostgreSQL features, Compare PostgreSQL
@subsubsection Benchmarking MySQL and PostgreSQL

@cindex PostgreSQL vs. MySQL, benchmarks

The only @code{Open Source} benchmark that we know of that can be used to
benchmark MySQL Server and PostgreSQL (and other databases) is our own. It can
be found at @uref{http://www.mysql.com/information/benchmarks.html}.

We have many times asked the PostgreSQL developers and some PostgreSQL
users to help us extend this benchmark to make it the definitive benchmark
for databases, but unfortunately we haven't gotten any feedback for this.

We, the MySQL developers, have, because of this, spent a lot of hours to get
maximum performance from PostgreSQL for the benchmarks, but because we
don't know PostgreSQL intimately, we are sure that there are things that
we have missed.  We have on the benchmark page documented exactly how we
did run the benchmark so that it should be easy for anyone to repeat and
verify our results.

The benchmarks are usually run with and without the @code{--fast} option.
When run with @code{--fast} we are trying to use every trick the server can
do to get the code to execute as fast as possible.  The idea is that the
normal run should show how the server would work in a default setup and
the @code{--fast} run shows how the server would do if the application
developer would use extensions in the server to make his application run
faster.

When running with PostgreSQL and @code{--fast} we do a @code{VACUUM}
after every major table @code{UPDATE} and @code{DROP TABLE} to make the
database in perfect shape for the following @code{SELECT}s.  The time for
@code{VACUUM} is measured separately.

When running with PostgreSQL 7.1.1 we could, however, not run with
@code{--fast} because during the @code{INSERT} test, the postmaster (the
PostgreSQL deamon) died and the database was so corrupted that it was
impossible to restart postmaster.   After this happened twice, we decided
to postpone the @code{--fast} test until the next PostgreSQL release.  The
details about the machine we run the benchmark on can be found on the
benchmark page.

Before going to the other benchmarks we know of, we would like to give
some background on benchmarks.

It's very easy to write a test that shows @strong{any} database to be the best
database in the world, by just restricting the test to something the
database is very good at and not testing anything that the database is
not good at.  If one, after doing this, summarises the result as
a single figure, things are even easier.

This would be like us measuring the speed of MySQL Server compared to PostgreSQL
by looking at the summary time of the MySQL benchmarks on our web page.
Based on this MySQL Server would be more than 40 times faster than PostgreSQL,
something that is, of course, not true.  We could make things even worse
by just taking the test where PostgreSQL performs worst and claim that
MySQL Server is more than 2000 times faster than PostgreSQL.

The case is that MySQL does a lot of optimisations that PostgreSQL
doesn't do. This is, of course, also true the other way around.  An SQL
optimiser is a very complex thing, and a company could spend years
just making the optimiser faster and faster.

When looking at the benchmark results you should look for things that
you do in your application and just use these results to decide which
database would be best suited for your application.  The benchmark
results also show things a particular database is not good at and should
give you a notion about things to avoid and what you may have to do in
other ways.

We know of two benchmark tests that claim that PostgreSQL performs better
than MySQL Server.  These both where multi-user tests, a test that we here at
MySQL AB haven't had time to write and include in the benchmark suite,
mainly because it's a big task to do this in a manner that is fair to
all databases.

One is the benchmark paid for by Great Bridge, the company that for 16 months
attempted to build a business based on PostgreSQL but now has ceased
operations. This is probably the worst benchmark we have ever seen anyone
conduct.  This was not only tuned to only test what PostgreSQL is absolutely
best at, but it was also totally unfair to every other database involved in
the test.

@strong{Note}:  We know that even some of the main PostgreSQL
developers did not like the way Great Bridge conducted the benchmark, so we
don't blame the PostgreSQL team for the way the benchmark was done.

This benchmark has been condemned in a lot of postings and newsgroups, so
here we will just briefly repeat some things that were wrong with it.

@itemize @bullet
@item
The tests were run with an expensive commercial tool that makes it
impossible for an @code{Open Source} company like us to verify the benchmarks,
or even check how the benchmarks were really done.  The tool is not even
a true benchmark tool, but an application/setup testing tool.  To refer
to this as a ``standard'' benchmark tool is to stretch the truth a long way.

@item
Great Bridge admitted that they had optimised the PostgreSQL database
(with @code{VACUUM} before the test) and tuned the startup for the tests,
something they hadn't done for any of the other databases involved.  They
say ``This process optimises indexes and frees up disk space a bit.  The
optimised indexes boost performance by some margin.''  Our benchmarks
clearly indicate that the difference in running a lot of selects on a
database with and without @code{VACUUM} can easily differ by a factor
of 10.

@item
The test results were also strange.  The AS3AP test documentation
mentions that the test does ``selections, simple joins, projections,
aggregates, one-tuple updates, and bulk updates.''

PostgreSQL is good at doing @code{SELECT}s and @code{JOIN}s (especially
after a @code{VACUUM}), but doesn't perform as well on @code{INSERT}s or
@code{UPDATE}s.  The benchmarks seem to indicate that only @code{SELECT}s
were done (or very few updates).  This could easily explain the good results
for PostgreSQL in this test.  The bad results for MySQL will be obvious a
bit down in this document.

@item
They did run the so-called benchmark from a Windows machine against a
Linux machine over ODBC, a setup that no normal database user would ever
do when running a heavy multi-user application.  This tested more the
ODBC driver and the Windows protocol used between the clients than the
database itself.

@item
When running the database against Oracle and MS-SQL (Great Bridge has
indirectly indicated the databases they used in the test), they
didn't use the native protocol but instead ODBC.  Anyone that has ever
used Oracle knows that all real applications use the native interface
instead of ODBC.  Doing a test through ODBC and claiming that the results
had anything to do with using the database in a real-world situation can't
be regarded as fair.  They should have done two tests with and without ODBC
to provide the right facts (after having gotten experts to tune all involved
databases, of course).

@item
They refer to the TPC-C tests, but they don't mention anywhere that the
test they did was not a true TPC-C test and they were not even allowed to
call it a TPC-C test.  A TPC-C test can only be conducted by the rules
approved by the TPC Council (@uref{http://www.tpc.org/}). Great Bridge
didn't do that.  By doing this they have both violated the TPC trademark
and miscredited their own benchmarks.  The rules set by the TPC Council
are very strict to ensure that no one can produce false results or make
unprovable statements.  Apparently Great Bridge wasn't interested in
doing this.

@item
After the first test, we contacted Great Bridge and mentioned to them
some of the obvious mistakes they had done with MySQL Server:

@itemize @minus
@item
Running with a debug version of our ODBC driver

@item
Running on a Linux system that wasn't optimised for threads

@item
Using an old MySQL version when there was a recommended newer one available

@item
Not starting MySQL Server with the right options for heavy multi-user use (the
default installation of MySQL Server is tuned for minimal resource use)
@end itemize

Great Bridge did run a new test, with our optimised ODBC driver and with
better startup options for MySQL Server, but refused to either use our updated
glibc library or our standard binary (used by 80% of our users), which was
statically linked with a fixed glibc library.

According to what we know, Great Bridge did nothing to ensure that the
other databases were set up correctly to run well in their test
environment.  We are sure, however, that they didn't contact Oracle or
Microsoft to ask for their advice in this matter. ;)

@item
The benchmark was paid for by Great Bridge, and they decided to publish
only partial, chosen results (instead of publishing it all).
@end itemize

Tim Perdue, a long-time PostgreSQL fan and a reluctant MySQL user,
published a comparison on PHPbuilder
(@uref{http://www.phpbuilder.com/columns/tim20001112.php3}).

When we became aware of the comparison, we phoned Tim Perdue about this
because there were a lot of strange things in his results.  For example,
he claimed that MySQL Server had a problem with five users in his tests, when we
know that there are users with similar machines as his that are using
MySQL Server with 2000 simultaneous connections doing 400 queries per second.
(In this case the limit was the web bandwidth, not the database.)

It sounded like he was using a Linux kernel that either had some
problems with many threads, such as kernels before 2.4, which had a problem
with many threads on multi-CPU machines. We have documented in this manual
how to fix this and Tim should be aware of this problem.

The other possible problem could have been an old glibc library and
that Tim didn't use a MySQL binary from our site, which is linked with
a corrected glibc library, but had compiled a version of his own.
In any of these cases, the symptom would have been exactly what Tim
had measured.

We asked Tim if we could get access to his data so that we could repeat
the benchmark and if he could check the MySQL version on the machine to
find out what was wrong and he promised to come back to us about this.
He has not done that yet.

Because of this we can't put any trust in this benchmark either. :(

Over time things also change and the preceding benchmarks are not that
relevant anymore.  MySQL Server now has a couple of different table handlers
with different speed/concurrency tradeoffs.  @xref{Table types}. It
would be interesting to see how the above tests would run with the
different transactional table types in MySQL Server.  PostgreSQL
has, of course,
also got new features since the test was made.  As these tests are
not publicly available there is no way for us to know how the
database would perform in the same tests today.


Conclusion:

The only benchmarks that exist today that anyone can download and run
against MySQL Server and PostgreSQL are the MySQL benchmarks.
We here at MySQL AB
believe that @code{Open Source} databases should be tested with @code{Open Source} tools!
This is the only way to ensure that no one does tests that nobody can
reproduce and use this to claim that one database is better than another.
Without knowing all the facts it's impossible to answer the claims of the
tester.

The thing we find strange is that every test we have seen about
PostgreSQL, that is impossible to reproduce, claims that PostgreSQL is
better in most cases while our tests, which anyone can reproduce,
clearly show otherwise.  With this we don't want to say that PostgreSQL
isn't good at many things (it is!) or that it isn't faster than MySQL Server
under certain conditions.  We would just like to see a fair test where
PostgreSQL performs very well, so that we could get some friendly
competition going!

For more information about our benchmark suite, see @ref{MySQL Benchmarks}.

We are working on an even better benchmark suite, including multi-user
tests, and a better documentation of what the individual tests really
do and how to add more tests to the suite.


@node Installing, Tutorial, Introduction, Top
@chapter MySQL Installation

@cindex installing, overview

@menu
* Quick Standard Installation::  Quick Standard Installation of MySQL
* General Installation Issues::  General Installation Issues
* Installing source::           Installing a MySQL source distribution
* Post-installation::           Post-installation setup and testing
* Upgrade::                     Upgrading/Downgrading MySQL
* Operating System Specific Notes::  Operating System Specific Notes
* Perl support::                Perl Installation Comments
@end menu

This chapter describes how to obtain and install MySQL:

@itemize @bullet
@item
For a list of sites from which you can obtain MySQL, see
@ref{Getting MySQL, , Getting MySQL}.

@item
To see which platforms are supported, see @ref{Which OS}. Please note that
not all supported systems are equally good for running MySQL on them.
On some it is much more robust and efficient than others@-see @ref{Which OS}
for details.

@item
Several versions of MySQL are available in both binary and
source distributions.  We also provide public access to our current
source tree for those who want to see our most recent developments and
help us test new code.  To determine which version and type of
distribution you should use, see @ref{Which version}. When in doubt,
use the binary distribution.

@item
Installation instructions for binary and source distributions are described
in @ref{Installing binary}, and @ref{Installing source}.  Each set of
instructions includes a section on system-specific problems you may run
into.

@item
For post-installation procedures, see @ref{Post-installation}.  These
procedures apply whether you install MySQL using a binary or
source distribution.
@end itemize


@node Quick Standard Installation, General Installation Issues, Installing, Installing
@section Quick Standard Installation of MySQL

@c This node name is special
@menu
* Linux-RPM::                   Installing MySQL on Linux
* Windows installation::        Installing MySQL on Windows
@end menu


@node Linux-RPM, Windows installation, Quick Standard Installation, Quick Standard Installation
@subsection Installing MySQL on Linux

@cindex RPM file
@cindex RedHat Package Manager

The recommended way to install MySQL on Linux is by using an RPM
file. The MySQL RPMs are currently being built on a RedHat Version
6.2 system but should work on other versions of Linux that support @code{rpm}
and use @code{glibc}.

If you have problems with an RPM file, for example, if you receive the error
``@code{Sorry, the host 'xxxx' could not be looked up}''@-see
@ref{Binary notes-Linux}.

The RPM files you may want to use are:

@itemize @bullet
@item @code{MySQL-VERSION.i386.rpm}

The MySQL server.  You will need this unless you only want to
connect to a MySQL server running on another machine.

@item @code{MySQL-client-VERSION.i386.rpm}

The standard MySQL client programs. You probably always want to
install this package.

@item @code{MySQL-bench-VERSION.i386.rpm}

Tests and benchmarks. Requires Perl and msql-mysql-modules RPMs.

@item @code{MySQL-devel-VERSION.i386.rpm}

Libraries and include files needed if you want to compile other
MySQL clients, such as the Perl modules.

@item @code{MySQL-VERSION.src.rpm}

This contains the source code for all of the previous packages. It can also
be used to try to build RPMs for other architectures (for example, Alpha
or SPARC).
@end itemize

To see all files in an RPM package, run:
@example
shell> rpm -qpl MySQL-VERSION.i386.rpm
@end example

To perform a standard minimal installation, run:

@example
shell> rpm -i MySQL-VERSION.i386.rpm MySQL-client-VERSION.i386.rpm
@end example

To install just the client package, run:

@example
shell> rpm -i MySQL-client-VERSION.i386.rpm
@end example

The RPM places data in @file{/var/lib/mysql}. The RPM also creates the
appropriate entries in @file{/etc/rc.d/} to start the server automatically
at boot time.  (This means that if you have performed a previous
installation, you may want to make a copy of your previously installed
MySQL startup file if you made any changes to it, so you don't lose
your changes.)

After installing the RPM file(s), the @code{mysqld} daemon should be running
and you should now be able to start using MySQL.
@xref{Post-installation}.

If something goes wrong, you can find more information in the binary
installation chapter. @xref{Installing binary}.


@node Windows installation,  , Linux-RPM, Quick Standard Installation
@subsection Installing MySQL on Windows

The MySQL server for Windows is available in two distribution types:
@enumerate
@item
The binary distribution contains a setup program which installs
everything you need so that you can start the server immediately.
@item
The source distribution contains all the code and support files
for building the executables using the VC++ 6.0 compiler.
@xref{Windows source build}.
@end enumerate

Generally speaking, you should use the binary distribution.

You will need the following:
@itemize @bullet
@item
A 32-bit Windows Operating System such as 9x, Me, NT, 2000, or XP.
The NT family (NT, Windows 2000 and XP) permits running the MySQL server
as a service. @xref{NT start}.

If you want to use tables bigger than 4G, you should install MySQL
on an NTFS or newer filesystem. Don't forget to use @code{MAX_ROWS} and
@code{AVG_ROW_LENGTH} when you create the table. @xref{CREATE TABLE}.
@item
TCP/IP protocol support.
@item
A copy of the MySQL binary or distribution for Windows, which
can be downloaded from @uref{http://www.mysql.com/downloads/}.

Note: The distribution files are supplied with a zipped format
and we recommend the use of an adequate FTP client with resume
feature to avoid corruption of files during the download process.
@item
A @code{ZIP} program to unpack the distribution file.
@item
Enough space on the hard drive to unpack, install, and create the
databases in accorandance with your requirements.
@item
If you plan to connect to the MySQL server via @code{ODBC}, you
will also need the @code{MyODBC} driver. @xref{ODBC}.
@end itemize


@menu
* Windows binary installation::  Installing the Binaries
* Windows prepare environment::  Preparing the Windows MySQL Environment
* Windows server first start::  Starting the Server for the First Time
@end menu

@node Windows binary installation, Windows prepare environment, Windows installation, Windows installation
@subsubsection Installing the Binaries

@enumerate
@item
If you are working on an NT/2000/XP server, logon as a user with
with administrator privileges.
@item
If you are doing an upgrade of an earlier MySQL installation,
it is necessary to stop the server. If you are running the server
as a service, use:

@example
C:\> NET STOP MySQL
@end example

Otherwise, use:

@example
C:\mysql\bin> mysqladmin -u root shutdown
@end example

@item
On NT/2000/XP machines, if you want to change the server executable
(e.g., -max or -nt), it is also necessary to remove the service:

@example
C:\mysql\bin> mysqld-max-nt --remove
@end example

@item
Unzip the distribution file to a temporary directory.
@item
Run the @file{setup.exe} file to begin the installation process.
If you want to install into another directory than the default
@file{c:\mysql}, use the @code{Browse} button to specify your
preferred directory.
@item
Finish the install process.
@end enumerate


@node Windows prepare environment, Windows server first start, Windows binary installation, Windows installation
@subsubsection Preparing the Windows MySQL Environment

Starting with MySQL 3.23.38, the Windows distribution includes
both the normal and the @strong{MySQL-Max} server binaries.
Here is a list of the different MySQL servers you can use:

@multitable @columnfractions .20 .80
@item @strong{Binary} @tab @strong{Description}
@item @code{mysqld} @tab
Compiled with full debugging and automatic memory allocation
checking, symbolic links, InnoDB, and BDB tables.
@item @code{mysqld-opt} @tab
Optimised binary with no support for transactional tables.
@item @code{mysqld-nt} @tab
Optimised binary for NT/2000/XP with support for named pipes.
You can run this version on Windows 9x/Me, but in this case no
named pipes are created and you must have TCP/IP installed.
@item @code{mysqld-max} @tab
Optimised binary with support for symbolic links,
InnoDB and BDB tables.
@item @code{mysqld-max-nt} @tab
Like @code{mysqld-max}, but compiled with support for named pipes.
@end multitable

Starting from 3.23.50, named pipes are only enabled if one starts mysqld with
@code{--enable-named-pipe}.

All of the preceding binaries are optimised for the Pentium Pro
processor but should work on any Intel processor >= i386.

You will need to use an option file to specify your MySQL configuration
under the following circumstances:
@itemize @bullet
@item
The installation or data directories are different from the default
locations (@file{c:\mysql} and @file{c:\mysql\data}).
@item
You want to use one of these servers:
@itemize @bullet
@item mysqld.exe
@item mysqld-max.exe
@item mysqld-max-nt.exe
@end itemize
@item
You need to tune the server settings.
@end itemize

Normally you can use the @code{WinMySQLAdmin} tool to edit the
option file @code{my.ini}.  In this case you don't have to worry
about the following section.

There are two option files with the same function: @file{my.cnf} and
@file{my.ini}.  However, to avoid confusion, it's best if you use only
of one them. Both files are plain text.  The @file{my.cnf} file, if used,
should be created in the root directory of the C drive.  The @file{my.ini}
file, if used, should be created in the Windows system directory.  (This
directory is typically something like @file{C:\WINDOWS} or @file{C:\WINNT}.
You can determine its exact location from the value of the @code{windir}
environment variable.)  MySQL looks first for the @code{my.ini} file,
then for the @file{my.cnf} file.

If your PC uses a boot loader where the C drive isn't the boot drive,
your only option is to use the @file{my.ini} file.  Also note that
if you use the @code{WinMySQLAdmin} tool, it uses only the @file{my.ini}
file.  The @file{\mysql\bin} directory contains a help file with
instructions for using this tool.

Using @code{notepad.exe}, create the option file and edit the
@code{[mysqld]} section to specify values for the @code{basedir} and
@code{datadir} parameters:

@example
[mysqld]
# set basedir to installation path, e.g., c:/mysql
basedir=the_install_path
# set datadir to location of data directory,
# e.g., c:/mysql/data or d:/mydata/data
datadir=the_data_path
@end example

Note that Windows pathnames should be specified in option files using
forward slashes rather than backslashes.  If you do use backslashes, you
must double them.

If you would like to use a data directory different from the default of
@file{c:\mysql\data}, you must copy the entire contents of the
@file{c:\mysql\data} directory to the new location.

If you want to use the @code{InnoDB} transactional tables, you
need to manually create two new directories to hold the InnoDB
data and log files@-e.g., @file{c:\ibdata} and @file{c:\iblogs}.
You will also need to add some extra lines to the option
file. @xref{InnoDB start}.

If you don't want to use @code{InnoDB} tables, add the
@code{skip-innodb} option to the option file.

Now you are ready to test starting the server.


@node Windows server first start,  , Windows prepare environment, Windows installation
@subsubsection Starting the Server for the First Time

Testing from a DOS command prompt is the best thing to do because
the server displays status messages that appear in the DOS window.
If something is wrong with your configuration, these messages will
make it easier for you to identify and fix any problems.

Make sure you are in the directory where the server is located, then
enter this command:

@example
C:\mysql\bin> mysqld-max --standalone
@end example

You should see the following messages as the server starts up:

@example
InnoDB: The first specified datafile c:\ibdata\ibdata1 did not exist:
InnoDB: a new database to be created!
InnoDB: Setting file c:\ibdata\ibdata1 size to 209715200
InnoDB: Database physically writes the file full: wait...
InnoDB: Log file c:\iblogs\ib_logfile0 did not exist: new to be created
InnoDB: Setting log file c:\iblogs\ib_logfile0 size to 31457280
InnoDB: Log file c:\iblogs\ib_logfile1 did not exist: new to be created
InnoDB: Setting log file c:\iblogs\ib_logfile1 size to 31457280
InnoDB: Log file c:\iblogs\ib_logfile2 did not exist: new to be created
InnoDB: Setting log file c:\iblogs\ib_logfile2 size to 31457280
InnoDB: Doublewrite buffer not found: creating new
InnoDB: Doublewrite buffer created
InnoDB: creating foreign key constraint system tables
InnoDB: foreign key constraint system tables created
011024 10:58:25  InnoDB: Started
@end  example

For further information about running MySQL on Windows,
see @ref{Windows}.


@node General Installation Issues, Installing source, Quick Standard Installation, Installing
@section General Installation Issues

@c @node Methods of Installation, , ,
@c @subsection Methods of Installation

@c FIX: this needs to be written?

@menu
* Getting MySQL::               How to Get MySQL
* Which OS::                    Operating Systems Supported by MySQL
* Which version::               Which MySQL Version to Use
* Installation layouts::        Installation Layouts
* Many versions::               How and When Updates Are Released
* MySQL binaries::              MySQL Binaries Compiled by MySQL AB
* Installing binary::           Installing a MySQL Binary Distribution
@end menu


@node Getting MySQL, Which OS, General Installation Issues, General Installation Issues
@subsection How to Get MySQL

@cindex downloading
@cindex MySQL version
@cindex version, latest
@cindex getting MySQL
@cindex mirror sites
@cindex URLS for downloading MySQL

Check the MySQL homepage (@uref{http://www.mysql.com/}) for
information about the current version and for downloading instructions.

Our main download mirror is located at:

@uref{http://mirrors.sunsite.dk/mysql/}

If you are interested in becoming a MySQL mirror site, you may
anonymously rsync with: @code{rsync://sunsite.dk/ftp/mirrors/mysql/}.  Please
send e-mail to @email{webmaster@@mysql.com} notifying us of your mirror to be
added to the following list.

If you have problems downloading from our main site, try using one of the
following mirrors.

Please report bad or out-of-date mirrors to @email{webmaster@@mysql.com}.

@c START_OF_MIRROR_LISTING

@c Mirrors list is created by PHP script (that really needs to be documented!) from database (tfr@mysql.com)
@include mirrors.texi

@c END_OF_MIRROR_LISTING


@node Which OS, Which version, Getting MySQL, General Installation Issues
@subsection Operating Systems Supported by MySQL

@cindex operating systems, supported
@cindex native thread support
@cindex thread support
@cindex process support
@cindex support, for operating systems

We use GNU Autoconf, so it is possible to port MySQL to all modern
systems with working Posix threads and a C++ compiler.  (To compile only the
client code, a C++ compiler is required but not threads.)  We use and develop
the software ourselves primarily on Sun Solaris (Versions 2.5 - 2.7) and
SuSE Linux Version 7.x.

Note that for many operating systems, the native thread support works only
in the latest versions. MySQL has been reported to compile
successfully on the following operating system/thread package combinations:

@itemize @bullet
@item
AIX 4.x with native threads.  @xref{IBM-AIX}.
@item
Amiga.
@item
BSDI 2.x with the included MIT-pthreads package.  @xref{BSDI}.
@item
BSDI 3.0, 3.1 and 4.x with native threads.  @xref{BSDI}.
@item
DEC Unix 4.x with native threads.  @xref{Alpha-DEC-UNIX}.
@item
FreeBSD 2.x with the included MIT-pthreads package.  @xref{FreeBSD}.
@item
FreeBSD 3.x and 4.x with native threads.  @xref{FreeBSD}.
@item
HP-UX 10.20 with the included MIT-pthreads package.  @xref{HP-UX 10.20}.
@item
HP-UX 11.x with the native threads.  @xref{HP-UX 11.x}.
@item
Linux 2.0+ with LinuxThreads 0.7.1+ or @code{glibc} 2.0.7+.  @xref{Linux}.
@item
Mac OS X Server.  @xref{Mac OS X}.
@item
NetBSD 1.3/1.4 Intel and NetBSD 1.3 Alpha (Requires GNU make).  @xref{NetBSD}.
@item
OpenBSD > 2.5 with native threads. OpenBSD < 2.5 with the included
MIT-pthreads package.  @xref{OpenBSD}.
@item
OS/2 Warp 3, FixPack 29 and OS/2 Warp 4, FixPack 4.  @xref{OS/2}.
@item
SGI Irix 6.x with native threads.  @xref{SGI-Irix}.
@item
Solaris 2.5 and above with native threads on SPARC and x86.  @xref{Solaris}.
@item
SunOS 4.x with the included MIT-pthreads package.  @xref{Solaris}.
@item
Caldera (SCO) OpenServer with a recent port of the FSU Pthreads package. 
@xref{Caldera}.
@item
Caldera (SCO) UnixWare 7.0.1.  @xref{Caldera Unixware}.
@item
Tru64 Unix
@item
Windows 9x, Me, NT, 2000 and XP.  @xref{Windows}.
@end itemize

Note that not all platforms are suited equally well for running
MySQL. How well a certain platform is suited for a high-load
mission-critical MySQL server is determined by the following
factors:

@itemize @bullet
@item
General stability of the thread library. A platform may have excellent
reputation otherwise, but if the thread library is unstable in the code
that is called by MySQL, even if everything else is perfect, MySQL will
be only as stable as the thread library.

@item
The ability of the kernel and/or thread library to take advantage of
@strong{SMP} on multi-processor systems. In other words, when a process
creates a thread, it should be possible for that thread to run on a different
CPU than the original process.

@item
The ability of the kernel and/or the thread library to run many threads which
acquire/release a mutex over a short critical region frequently without
excessive context switches. In other words, if the implementation of
@code{pthread_mutex_lock()} is too anxious to yield CPU time, this will hurt
MySQL tremendously. If this issue is not taken care of, adding extra CPUs
will actually make MySQL slower.

@item
General filesystem stability/performance.

@item
Ability of the filesystem to deal with large files at all and deal with them
efficiently, if your tables are big.

@item
Our level of expertise here at MySQL AB with the platform. If we know
a platform well, we introduce platform-specific optimisations/fixes enabled at
compile time. We can also provide advice on configuring your system optimally
for MySQL.

@item
The amount of testing of similar configurations we have done internally.

@item
The number of users that have successfully run MySQL on that
platform in similar configurations. If this number is high, the chances of
hitting some platform-specific surprises are much smaller.
@end itemize

Based on the preceding criteria, the best platforms for running
MySQL at this point are x86 with SuSE Linux 7.1, 2.4 kernel, and
ReiserFS (or any similar Linux distribution) and SPARC with Solaris 2.7
or 2.8. FreeBSD comes third, but we really hope it will join the top
club once the thread library is improved. We also hope that at some
point we will be able to include all other platforms on which
MySQL compiles, runs okay, but not quite with the same level of
stability and performance, into the top category. This will require some
effort on our part in cooperation with the developers of the OS/library
components MySQL depends upon. If you are interested in making
one of those components better, are in a position to influence their
development, and need more detailed instructions on what MySQL
needs to run better, send an e-mail to
@email{internals@@lists.mysql.com}.

Please note that the preceding comparison is not to say that one OS is better or
worse than the other in general. We are talking about choosing a particular OS
for a dedicated purpose@-running MySQL, and compare platforms in that
regard only. With this in mind, the result of this comparison
would be different if we included more issues into it. And in some cases,
the reason one OS is better than the other could simply be that we have put
forth more effort into testing on and optimising for that particular platform.
We are just stating our observations to help you decide on which
platform to use MySQL on in your setup.


@node Which version, Installation layouts, Which OS, General Installation Issues
@subsection Which MySQL Version to Use

@cindex MySQL binary distribution
@cindex MySQL source distribution
@cindex release numbers
@cindex version, choosing
@cindex choosing, a MySQL version

The first decision to make is whether you want to use the latest development
release or the last stable release:

@itemize @bullet
@item
Normally, if you are beginning to use MySQL for the first time or trying
to port it to some system for which there is no binary distribution, we
recommend going with the stable release (currently version 3.23).  Note
that all MySQL releases are checked with the MySQL benchmarks and an
extensive test suite before each release (even the development
releases).

@item
Otherwise, if you are running an old system and want to upgrade, but
don't want to take chances with a non-seamless upgrade, you should
upgrade to the latest in the same branch you are using (where only the
last version number is newer than yours).  We have tried to fix only
fatal bugs and make small, relatively safe changes to that version.
@end itemize

The second decision to make is whether you want to use a source
distribution or a binary distribution.  In most cases you should probably
use a binary distribution, if one exists for your platform, as this
generally will be easier to install than a source distribution.

In the following cases you probably will be better off with a source
installation:

@itemize @bullet
@item
If you want to install MySQL at some explicit location.  (The standard
binary distributions are ``ready to run'' at any place, but you may want
to get even more flexibility).

@item
To be able to satisfy different user requirements, we are providing two
different binary versions: one compiled with the non-transactional table
handlers (a small, fast binary), and one configured with the most
important extended options like transaction-safe tables.  Both versions
are compiled from the same source distribution.  All native @code{MySQL}
clients can connect to both MySQL versions.

The extended MySQL binary distribution is marked with the
@code{-max} suffix and is configured with the same options as
@code{mysqld-max}. @xref{mysqld-max, , @code{mysqld-max}}.

If you want to use the @code{MySQL-Max} RPM, you must first
install the standard @code{MySQL} RPM.

@item
If you want to configure @code{mysqld} with some extra features that are
not in the standard binary distributions.  Here is a list of the most
common extra options that you may want to use:

@itemize @bullet
@item @code{--with-innodb}
@item @code{--with-berkeley-db}
@item @code{--with-raid}
@item @code{--with-libwrap}
@item @code{--with-named-z-lib  (This is done for some of the binaries)}
@item @code{--with-debug[=full]}
@end itemize

@item
The default binary distribution is normally compiled with support
for all character sets and should work on a variety of processors from
the same processor family.

If you want a faster MySQL server you may want to recompile it
with support for only the character sets you need, use a better compiler
(like @code{pgcc}), or use compiler options that are better optimised for your
processor.

@item
If you have found a bug and reported it to the MySQL
development team you will probably receive a patch that you need to apply to
the source distribution to get the bug fixed.

@item
If you want to read (and/or modify) the C and C++ code that makes up
MySQL, you should get a source distribution.  The source code is
always the ultimate manual.  Source distributions also contain more
tests and examples than binary distributions.
@end itemize

@cindex naming, releases of MySQL
@cindex releases, naming scheme

The MySQL naming scheme uses release numbers that consist of three
numbers and a suffix.  For example, a release name like
@code{mysql-3.21.17-beta} is interpreted like this:

@itemize @bullet
@item
The first number (@code{3}) describes the file format.  All Version 3
releases have the same file format.

@item
The second number (@code{21}) is the release level.  Normally there are two to
choose from. One is the release/stable branch (currently @code{23}) and the
other is the development branch (currently @code{4.0}).  Normally both are
stable, but the development version may have quirks, may be missing documentation on
new features, or may fail to compile on some systems.

@item
The third number (@code{17}) is the version number within the
release level.  This is incremented for each new distribution.  Usually you
want the latest version for the release level you have chosen.

@item
The suffix (@code{beta}) indicates the stability level of the release.
The possible suffixes are:

@itemize @minus
@item
@code{alpha} indicates that the release contains some large section of
new code that hasn't been 100% tested.  Known bugs (usually there are none)
should be documented in the News section.  @xref{News}.  There are also new
commands and extensions in most alpha releases.  Active development that
may involve major code changes can occur on an alpha release, but everything
will be tested before doing a release.  There should be no known bugs in any
MySQL release.

@item
@code{beta} means that all new code has been tested.  No major new
features that could cause corruption on old code are added.  There should
be no known bugs.  A version changes from alpha to beta when there
haven't been any reported fatal bugs within an alpha version for at least
a month and we don't plan to add any features that could make any old command
more unreliable.

@item
@code{gamma} is a beta that has been around a while and seems to work fine.
Only minor fixes are added.  This is what many other companies call a release.

@item
If there is no suffix, it means that the version has been run for a
while at many different sites with no reports of bugs other than
platform-specific bugs.  Only critical bug fixes are applied to the
release. This is what we call a stable release.
@end itemize
@end itemize

All versions of MySQL are run through our standard tests and
benchmarks to ensure that they are relatively safe to use.  Because the
standard tests are extended over time to check for all previously found bugs,
the test suite keeps getting better.

@cindex releases, testing
@cindex testing, of MySQL releases

Note that all releases have been tested at least with:

@table @asis
@item An internal test suite
This is part of a production system for a customer.  It has many tables with
hundreds of megabytes of data.

@item The MySQL benchmark suite
This runs a range of common queries.  It is also a test to see whether the
latest batch of optimisations actually made the code faster.
@xref{MySQL Benchmarks}.

@item The @code{crash-me} test
This tries to determine what features the database supports and what its
capabilities and limitations are.  @xref{MySQL Benchmarks}.
@end table

Another test is that we use the newest MySQL version in our internal
production environment, on at least one machine.  We have more than 100
gigabytes of data to work with.


@node Installation layouts, Many versions, Which version, General Installation Issues
@subsection Installation Layouts

@cindex installation layouts
@cindex layout of installation
@cindex directory structure, default
@cindex default installation location

This section describes the default layout of the directories created by
installing binary and source distributions.

A binary distribution is installed by unpacking it at the installation
location you choose (typically @file{/usr/local/mysql}) and creates the
following directories in that location:

@multitable @columnfractions .15 .45
@item @strong{Directory} @tab @strong{Contents of directory}
@item @file{bin}         @tab Client programs and the @code{mysqld} server
@item @file{data}        @tab Log files, databases
@item @file{include}     @tab Include (header) files
@item @file{lib}         @tab Libraries
@item @file{scripts}     @tab @code{mysql_install_db}
@item @file{share/mysql} @tab Error message files
@item @file{sql-bench}   @tab Benchmarks
@end multitable

A source distribution is installed after you configure and compile it.  By
default, the installation step installs files under @file{/usr/local}, in the
following subdirectories:

@multitable @columnfractions .15 .45
@item @strong{Directory} @tab @strong{Contents of directory}
@item @file{bin}         @tab Client programs and scripts
@item @file{include/mysql}  @tab Include (header) files
@item @file{info}	 @tab Documentation in Info format
@item @file{lib/mysql}   @tab Libraries
@item @file{libexec}     @tab The @code{mysqld} server
@item @file{share/mysql} @tab Error message files
@item @file{sql-bench}   @tab Benchmarks and @code{crash-me} test
@item @file{var}         @tab Databases and log files
@end multitable

Within an installation directory, the layout of a source installation differs
from that of a binary installation in the following ways:

@itemize @bullet
@item
The @code{mysqld} server is installed in the @file{libexec}
directory rather than in the @file{bin} directory.

@item
The data directory is @file{var} rather than @file{data}.

@item
@code{mysql_install_db} is installed in the @file{/usr/local/bin} directory
rather than in @file{/usr/local/mysql/scripts}.

@item
The header file and library directories are @file{include/mysql} and
@file{lib/mysql} rather than @file{include} and @file{lib}.
@end itemize

You can create your own binary installation from a compiled source
distribution by executing the script @file{scripts/make_binary_distribution}.


@node Many versions, MySQL binaries, Installation layouts, General Installation Issues
@subsection How and When Updates Are Released

@cindex releases, updating
@cindex updating, releases of MySQL

MySQL is evolving quite rapidly here at MySQL AB and we
want to share this with other MySQL users.  We try to make a release
when we have very useful features that others seem to have a need for.

We also try to help out users who request features that are easy to
implement.  We take note of what our licensed users want to have, and
we especially take note of what our extended e-mail supported customers
want and try to help them out.

No one has to download a new release.  The News section will tell you if
the new release has something you really want.  @xref{News}.

We use the following policy when updating MySQL:

@itemize @bullet
@item
For each minor update, the last number in the version string is incremented.
When there are major new features or minor incompatibilities with previous
versions, the second number in the version string is incremented.  When the
file format changes, the first number is increased.

@item
Stable-tested releases are meant to appear about 1-2 times a year, but
if small bugs are found, a release with only bug fixes will be released.

@item
Working releases/bug fixes to old releases are meant to appear about
every 1-8 weeks.

@item
Binary distributions for some platforms will be made by us for major releases.
Other people may make binary distributions for other systems but probably
less frequently.

@item
We usually make patches available as soon as we have located and fixed
small bugs. They are posted to @email{bugs@@lists.mysql.com} and will
be added to the next release.

@item
For non-critical but annoying bugs, we will add them the MySQL source
repository and they will be fixed in the next release.

@item
If there is, by any chance, a fatal bug in a release we will make a new
release as soon as possible.  We would like other companies to do this,
too.
@end itemize

The current stable release is Version 3.23; we have already moved active
development to Version 4.0.  Bugs will still be fixed in the stable version.
We don't believe in a complete freeze, as this also leaves out bug fixes
and things that ``must be done.''  ``Somewhat frozen'' means that we may
add small things that ``almost surely will not affect anything that's
already working.''

MySQL uses a slightly different naming scheme from most other products.
In general it's relatively safe to use any version that has been out for
a couple of weeks without being replaced with a new version.
@xref{Which version}.

@node MySQL binaries, Installing binary, Many versions, General Installation Issues
@subsection MySQL Binaries Compiled by MySQL AB

@cindex binary distributions

As a service, we at MySQL AB provide a set of binary distributions
of MySQL that are compiled at our site or at sites where customers
kindly have given us access to their machines.

These distributions are generated with @code{scripts/make_binary_distribution}
and are configured with the following compilers and options:

@table @asis
@item SunOS 4.1.4 2 sun4c with @code{gcc} 2.7.2.1
@code{CC=gcc CXX=gcc CXXFLAGS="-O3 -felide-constructors" ./configure --prefix=/usr/local/mysql --disable-shared --with-extra-charsets=complex --enable-assembler}

@item SunOS 5.5.1 (and above) sun4u with @code{egcs} 1.0.3a or 2.90.27 or gcc 2.95.2 and newer
@code{CC=gcc CFLAGS="-O3" CXX=gcc CXXFLAGS="-O3 -felide-constructors -fno-exceptions -fno-rtti" ./configure --prefix=/usr/local/mysql --with-low-memory --with-extra-charsets=complex --enable-assembler}

@item SunOS 5.6 i86pc with @code{gcc} 2.8.1
@code{CC=gcc CXX=gcc CXXFLAGS=-O3 ./configure --prefix=/usr/local/mysql --with-low-memory --with-extra-charsets=complex}

@item Solaris 2.8 sparc with @code{gcc 2.95.3}
@code{CC=gcc CFLAGS="-O3 -fno-omit-frame-pointer" CXX=gcc CXXFLAGS="-O3 -fno-omit-frame-pointer -felide-constructors -fno-exceptions -fno-rtti" ./configure --prefix=/usr/local/mysql "--with-comment=Official MySQL binary" --with-extra-charsets=complex "--with-server-suffix=" --enable-thread-safe-client --enable-local-infile --enable-assembler --disable-shared}

@item Linux 2.0.33 i386 with @code{pgcc} 2.90.29 (@code{egcs} 1.0.3a)
@code{CFLAGS="-O3 -mpentium -mstack-align-double" CXX=gcc CXXFLAGS="-O3 -mpentium -mstack-align-double -felide-constructors -fno-exceptions -fno-rtti" ./configure --prefix=/usr/local/mysql --enable-assembler --with-mysqld-ldflags=-all-static --with-extra-charsets=complex}

@item Linux 2.2.x with x686 with @code{gcc} 2.95.2
@code{CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro -felide-constructors -fno-exceptions -fno-rtti" ./configure --prefix=/usr/local/mysql --enable-assembler --with-mysqld-ldflags=-all-static --disable-shared --with-extra-charset=complex}

@item SCO 3.2v5.0.4 i386 with @code{gcc} 2.7-95q4
@code{CC=gcc CXX=gcc CXXFLAGS=-O3 ./configure --prefix=/usr/local/mysql --with-extra-charsets=complex}

@item AIX 2 4 with @code{gcc} 2.7.2.2
@code{CC=gcc CXX=gcc CXXFLAGS=-O3 ./configure --prefix=/usr/local/mysql --with-extra-charsets=complex}

@item OSF/1 V4.0 564 alpha with @code{gcc} 2.8.1
@code{CC=gcc CFLAGS=-O CXX=gcc CXXFLAGS=-O3 ./configure --prefix=/usr/local/mysql --with-low-memory --with-extra-charsets=complex}

@item Irix 6.3 IP32 with @code{gcc} 2.8.0
@code{CC=gcc CXX=gcc CXXFLAGS=-O3 ./configure --prefix=/usr/local/mysql --with-extra-charsets=complex}

@item BSDI BSD/OS 3.1 i386 with @code{gcc} 2.7.2.1
@code{CC=gcc CXX=gcc CXXFLAGS=-O ./configure --prefix=/usr/local/mysql --with-extra-charsets=complex}

@item BSDI BSD/OS 2.1 i386 with @code{gcc} 2.7.2
@code{CC=gcc CXX=gcc CXXFLAGS=-O3 ./configure --prefix=/usr/local/mysql --with-extra-charsets=complex}

@item FreeBSD 4.4-stable i386 with @code{gcc} 2.95.3
@code{CC=gcc CFLAGS="-O3 -fno-omit-frame-pointer" CXX=gcc CXXFLAGS="-O3 -fno-omit-frame-pointer -felide-constructors -fno-exceptions -fno-rtti" ./configure --prefix=/usr/local/mysql "--with-comment=Official MySQL binary" --with-extra-charsets=complex "--with-server-suffix=" --enable-thread-safe-client --enable-local-infile --enable-assembler --with-named-z-libs=not-used --disable-shared}
@end table

Anyone who has more optimal options for any of the preceding
configurations listed can always mail them to the developer's mailing list at
@email{internals@@lists.mysql.com}.

RPM distributions prior to MySQL Version 3.22 are user-contributed.
Beginning with Version 3.22, the RPMs are generated by us at
MySQL AB.

If you want to compile a debug version of MySQL, you should add
@code{--with-debug} or @code{--with-debug=full} to the preceding configure lines
and remove any @code{-fomit-frame-pointer} options.

For the Windows distribution, please see @ref{Windows installation}.


@node Installing binary,  , MySQL binaries, General Installation Issues
@subsection Installing a MySQL Binary Distribution

@cindex installing, binary distribution
@cindex binary distributions, installing

See also @ref{Windows binary installation},
@ref{Linux-RPM}, and @ref{Building clients}.

You need the following tools to install a MySQL binary distribution:

@itemize @bullet
@item
GNU @code{gunzip} to uncompress the distribution.

@item
A reasonable @code{tar} to unpack the distribution. GNU @code{tar} is
known to work.  Sun @code{tar} is known to have problems.
@end itemize

@cindex RPM, defined
@cindex RedHat Package Manager
An alternative installation method under Linux is to use RPM (RedHat Package
Manager) distributions.  @xref{Linux-RPM}.

@c texi2html fails to split chapters if I use strong for all of this.
If you run into problems, @strong{please always use @code{mysqlbug}} when
posting questions to @email{mysql@@lists.mysql.com}.  Even if the problem
isn't a bug, @code{mysqlbug} gathers system information that will help others
solve your problem.  By not using @code{mysqlbug}, you lessen the likelihood
of getting a solution to your problem!  You will find @code{mysqlbug} in the
@file{bin} directory after you unpack the distribution.  @xref{Bug reports}.

@cindex commands, for binary distribution
The basic commands you must execute to install and use a MySQL
binary distribution are:

@example
shell> groupadd mysql
shell> useradd -g mysql mysql
shell> cd /usr/local
shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf -
shell> ln -s full-path-to-mysql-VERSION-OS mysql
shell> cd mysql
shell> scripts/mysql_install_db
shell> chown -R root  .
shell> chown -R mysql data
shell> chgrp -R mysql .
shell> bin/safe_mysqld --user=mysql &
or
shell> bin/mysqld_safe --user=mysql &
if you are running MySQL 4.x
@end example

@cindex adding, new users
@cindex new users, adding
@cindex users, adding

You can add new users using the @code{bin/mysql_setpermission} script if
you install the @code{DBI} and @code{Msql-Mysql-modules} Perl modules.

A more detailed description follows.

To install a binary distribution, follow these steps, then proceed
to @ref{Post-installation}, for post-installation setup and testing:

@enumerate
@item
Pick the directory under which you want to unpack the distribution, and move
into it.  In the following example, we unpack the distribution under
@file{/usr/local} and create a directory @file{/usr/local/mysql} into which
MySQL is installed.  (The following instructions, therefore, assume
you have permission to create files in @file{/usr/local}.  If that directory
is protected, you will need to perform the installation as @code{root}.)

@item
Obtain a distribution file from one of the sites listed in
@ref{Getting MySQL, , Getting MySQL}.

MySQL binary distributions are provided as compressed @code{tar}
archives and have names like @file{mysql-VERSION-OS.tar.gz}, where
@code{VERSION} is a number (for example, @code{3.21.15}), and @code{OS}
indicates the type of operating system for which the distribution is intended
(for example, @code{pc-linux-gnu-i586}).

@item
If you see a binary distribution marked with the @code{-max} prefix, this
means that the binary has support for transaction-safe tables and other
features. @xref{mysqld-max, , @code{mysqld-max}}.  Note that all binaries
are built from the same MySQL source distribution.

@item
Add a user and group for @code{mysqld} to run as:

@example
shell> groupadd mysql
shell> useradd -g mysql mysql
@end example

These commands add the @code{mysql} group and the @code{mysql} user.  The
syntax for @code{useradd} and @code{groupadd} may differ slightly on different
versions of Unix.  They may also be called @code{adduser} and @code{addgroup}.
You may wish to call the user and group something else instead of @code{mysql}.

@item
Change into the intended installation directory:

@example
shell> cd /usr/local
@end example

@item
Unpack the distribution and create the installation directory:

@example
shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf -
shell> ln -s full-path-to-mysql-VERSION-OS mysql
@end example

The first command creates a directory named @file{mysql-VERSION-OS}.  The
second command makes a symbolic link to that directory.  This lets you refer
more easily to the installation directory as @file{/usr/local/mysql}.

@item
Change into the installation directory:

@example
shell> cd mysql
@end example

You will find several files and subdirectories in the @code{mysql} directory.
The most important for installation purposes are the @file{bin} and
@file{scripts} subdirectories.

@table @file
@item bin
@tindex PATH environment variable
@tindex environment variable, PATH
This directory contains client programs and the server
You should add the full pathname of this directory to your
@code{PATH} environment variable so that your shell finds the MySQL
programs properly. @xref{Environment variables}.

@item scripts
This directory contains the @code{mysql_install_db} script used to initialise
the @code{mysql} database containing the grant tables that store the server
access permissions.
@end table

@item
If you would like to use @code{mysqlaccess} and have the MySQL
distribution in some non-standard place, you must change the location where
@code{mysqlaccess} expects to find the @code{mysql} client.  Edit the
@file{bin/mysqlaccess} script at approximately line 18.  Search for a line
that looks like this:

@example
$MYSQL     = '/usr/local/bin/mysql';    # path to mysql executable
@end example

Change the path to reflect the location where @code{mysql} actually is
stored on your system.  If you do not do this, you will get a @code{Broken
pipe} error when you run @code{mysqlaccess}.

@item
Create the MySQL grant tables (necessary only if you haven't
installed MySQL before):
@example
shell> scripts/mysql_install_db
@end example

Note that MySQL versions older than Version 3.22.10 started the
MySQL server when you run @code{mysql_install_db}.  This is no
longer true!

@item
Change ownership of binaries to @code{root} and ownership of the data
directory to the user that you will run @code{mysqld} as:

@example
shell> chown -R root  /usr/local/mysql/.
shell> chown -R mysql /usr/local/mysql/data
shell> chgrp -R mysql /usr/local/mysql/.
@end example

The first command changes the @code{owner} attribute of the files to the
@code{root} user, the second one changes the @code{owner} attribute of the
data directory to the @code{mysql} user, and the third one changes the
@code{group} attribute to the @code{mysql} group.

@item
If you want to install support for the Perl @code{DBI}/@code{DBD} interface,
see @ref{Perl support}.

@item
If you would like MySQL to start automatically when you boot your
machine, you can copy @code{support-files/mysql.server} to the location where
your system has its startup files.  More information can be found in the
@code{support-files/mysql.server} script itself and in
@ref{Automatic start}.

@end enumerate

After everything has been unpacked and installed, you should initialise
and test your distribution.

You can start the MySQL server with the following command:

@example
shell> bin/safe_mysqld --user=mysql &
@end example

Now proceed to @ref{safe_mysqld, , @code{safe_mysqld}}, and
@xref{Post-installation}.


@node Installing source, Post-installation, General Installation Issues, Installing
@section Installing a MySQL Source Distribution

@cindex installing, source distribution
@cindex source distribution, installing
@cindex installation overview

Before you proceed with the source installation, check first to see if our
binary is available for your platform and if it will work for you. We
put a lot of effort into making sure that our binaries are built with the
best possible options.

You need the following tools to build and install MySQL from source:

@itemize @bullet
@item
GNU @code{gunzip} to uncompress the distribution.

@item
A reasonable @code{tar} to unpack the distribution.  GNU @code{tar} is
known to work.  Sun @code{tar} is known to have problems.

@item
A working ANSI C++ compiler.  @code{gcc} >= 2.95.2, @code{egcs} >= 1.0.2
or @code{egcs 2.91.66}, SGI C++, and SunPro C++ are some of the
compilers that are known to work.  @code{libg++} is not needed when
using @code{gcc}.  @code{gcc} 2.7.x has a bug that makes it impossible
to compile some perfectly legal C++ files, such as
@file{sql/sql_base.cc}.  If you only have @code{gcc} 2.7.x, you must
upgrade your @code{gcc} to be able to compile MySQL. @code{gcc}
2.8.1 is also known to have problems on some platforms, so it should be
avoided if a new compiler exists for the platform.

@code{gcc} >= 2.95.2 is recommended when compiling MySQL
Version 3.23.x.

@item
A good @code{make} program.  GNU @code{make} is always recommended and is
sometimes required.  If you have problems, we recommend trying GNU
@code{make} 3.75 or newer.
@end itemize

If you are using a recent version of @code{gcc}, recent enough to understand the
@code{-fno-exceptions} option, it is @strong{very important} that you use
it. Otherwise, you may compile a binary that crashes randomly. We also
recommend that you use @code{-felide-constructors} and @code{-fno-rtti} along
with @code{-fno-exceptions}. When in doubt, do the following:

@example

CFLAGS="-O3" CXX=gcc CXXFLAGS="-O3 -felide-constructors -fno-exceptions \
       -fno-rtti" ./configure --prefix=/usr/local/mysql --enable-assembler \
       --with-mysqld-ldflags=-all-static

@end example

On most systems this will give you a fast and stable binary.

@c texi2html fails to split chapters if I use strong for all of this.
If you run into problems, @strong{please always use @code{mysqlbug}} when
posting questions to @email{mysql@@lists.mysql.com}.  Even if the problem
isn't a bug, @code{mysqlbug} gathers system information that will help others
solve your problem.  By not using @code{mysqlbug}, you lessen the likelihood
of getting a solution to your problem!  You will find @code{mysqlbug} in the
@file{scripts} directory after you unpack the distribution.
@xref{Bug reports}.

@menu
* Quick install::               Quick installation overview
* Applying patches::            Applying patches
* configure options::           Typical @code{configure} options
* Installing source tree::      Installing from the Development Source Tree
* Compilation problems::        Problems Compiling?
* MIT-pthreads::                MIT-pthreads Notes
* Windows source build::        Windows Source Distribution
@end menu


@node Quick install, Applying patches, Installing source, Installing source
@subsection Quick Installation Overview

The basic commands you must execute to install a MySQL source
distribution are:

@example
shell> groupadd mysql
shell> useradd -g mysql mysql
shell> gunzip < mysql-VERSION.tar.gz | tar -xvf -
shell> cd mysql-VERSION
shell> ./configure --prefix=/usr/local/mysql
shell> make
shell> make install
shell> scripts/mysql_install_db
shell> chown -R root  /usr/local/mysql
shell> chown -R mysql /usr/local/mysql/var
shell> chgrp -R mysql /usr/local/mysql
shell> cp support-files/my-medium.cnf /etc/my.cnf
shell> /usr/local/mysql/bin/safe_mysqld --user=mysql &
or
shell> /usr/local/mysql/bin/mysqld_safe --user=mysql &
if you are running MySQL 4.x.
@end example

If you want to have support for InnoDB tables, you should edit the
@code{/etc/my.cnf} file and remove the @code{#} character before the
parameter that starts with @code{innodb_...}.
@xref{Option files}, and @ref{InnoDB start}.

If you start from a source RPM, do the following:

@example
shell> rpm --rebuild MySQL-VERSION.src.rpm
@end example

This will make a binary RPM that you can install.

You can add new users using the @code{bin/mysql_setpermission} script if
you install the @code{DBI} and @code{Msql-Mysql-modules} Perl modules.

A more detailed description follows.

To install a source distribution, follow these steps, then proceed
to @ref{Post-installation}, for post-installation initialisation and testing:

@enumerate
@item
Pick the directory under which you want to unpack the distribution, and move
into it.

@item
Obtain a distribution file from one of the sites listed in
@ref{Getting MySQL, , Getting MySQL}.

@item
If you are interested in using Berkeley DB tables with MySQL, you
will need to obtain a patched version of the Berkeley DB source
code.  Please read the chapter on Berkeley DB tables before
proceeding.  @xref{BDB}.

MySQL source distributions are provided as compressed @code{tar}
archives and have names like @file{mysql-VERSION.tar.gz}, where
@code{VERSION} is a number like @value{mysql_version}.

@item
Add a user and group for @code{mysqld} to run as:

@example
shell> groupadd mysql
shell> useradd -g mysql mysql
@end example

These commands add the @code{mysql} group and the @code{mysql} user.  The
syntax for @code{useradd} and @code{groupadd} may differ slightly on different
versions of Unix.  They may also be called @code{adduser} and @code{addgroup}.
You may wish to call the user and group something else instead of @code{mysql}.

@item
Unpack the distribution into the current directory:
@example
shell> gunzip < /path/to/mysql-VERSION.tar.gz | tar xvf -
@end example

This command creates a directory named @file{mysql-VERSION}.

@item
Change into the top-level directory of the unpacked distribution:

@example
shell> cd mysql-VERSION
@end example

Note that currently you must configure and build MySQL from
this top-level directory.  You cannot build it in a different
directory.

@item
Configure the release and compile everything:

@example
shell> ./configure --prefix=/usr/local/mysql
shell> make
@end example

When you run @code{configure}, you might want to specify some options.
Run @code{./configure --help} for a list of options.
@ref{configure options, , @code{configure} options}, discusses some of the
more useful options.

If @code{configure} fails, and you are going to send mail to
@email{mysql@@lists.mysql.com} to ask for assistance, please include any
lines from @file{config.log} that you think can help solve the problem.  Also
include the last couple of lines of output from @code{configure} if
@code{configure} aborts.  Post the bug report using the @code{mysqlbug}
script.  @xref{Bug reports}.

If the compile fails, see @ref{Compilation problems}, for help with
a number of common problems.

@item
Install everything:

@example
shell> make install
@end example

You might need to run this command as @code{root}.

@item
Create the MySQL grant tables (necessary only if you haven't
installed MySQL before):

@example
shell> scripts/mysql_install_db
@end example

Note that MySQL versions older than Version 3.22.10 started the
MySQL server when you run @code{mysql_install_db}.  This is no
longer true!

@item
Change ownership of binaries to @code{root} and ownership of the data
directory to the user that you will run @code{mysqld} as:

@example
shell> chown -R root  /usr/local/mysql
shell> chown -R mysql /usr/local/mysql/var
shell> chgrp -R mysql /usr/local/mysql
@end example

The first command changes the @code{owner} attribute of the files to the
@code{root} user, the second one changes the @code{owner} attribute of the
data directory to the @code{mysql} user, and the third one changes the
@code{group} attribute to the @code{mysql} group.

@item
If you want to install support for the Perl @code{DBI}/@code{DBD} interface,
see @ref{Perl support}.

@item
If you would like MySQL to start automatically when you boot your
machine, you can copy @code{support-files/mysql.server} to the location where
your system has its startup files.  More information can be found in the
@code{support-files/mysql.server} script itself and in
@ref{Automatic start}.
@end enumerate

After everything has been installed, you should initialise and test your
distribution:

@example
shell> /usr/local/mysql/bin/safe_mysqld --user=mysql &
@end example

If that command fails immediately with @code{mysqld daemon ended}, you can
find some information in the file @file{mysql-data-directory/'hostname'.err}.
The likely reason is that you already have another @code{mysqld} server
running.  @xref{Multiple servers}.

Now proceed to @ref{Post-installation}.


@node Applying patches, configure options, Quick install, Installing source
@subsection Applying Patches

@cindex patches, applying
@cindex applying, patches

Sometimes patches appear on the mailing list or are placed in the
patches area of the MySQL web site
(@uref{http://www.mysql.com/Downloads/Patches/}).

To apply a patch from the mailing list, save the message in which the patch
appears in a file, change into the top-level directory of your MySQL
source tree, and run these commands:

@example
shell> patch -p1 < patch-file-name
shell> rm config.cache
shell> make clean
@end example

Patches from the FTP site are distributed as plain text files or as files
compressed with @code{gzip}.  Apply a plain patch as shown
previously for
mailing list patches.  To apply a compressed patch, change into the
top-level directory of your MySQL source tree and run these
commands:

@example
shell> gunzip < patch-file-name.gz | patch -p1
shell> rm config.cache
shell> make clean
@end example

After applying a patch, follow the instructions for a normal source install,
beginning with the @code{./configure} step.  After running the @code{make
install} step, restart your MySQL server.

You may need to bring down any currently running server before you run
@code{make install}.  (Use @code{mysqladmin shutdown} to do this.)  Some
systems do not allow you to install a new version of a program if it replaces
the version that is currently executing.


@node configure options, Installing source tree, Applying patches, Installing source
@subsection Typical @code{configure} Options

@findex without-server option
@cindex @code{configure} script
@cindex options, configure
@cindex configuration options
@cindex log files
@cindex files, log

The @code{configure} script gives you a great deal of control over how
you configure your MySQL distribution.  Typically you do this
using options on the @code{configure} command-line.  You can also affect
@code{configure} using certain environment variables.  @xref{Environment
variables}.  For a list of options supported by @code{configure}, run
this command:

@example
shell> ./configure --help
@end example

Some of the more commonly-used @code{configure} options are described here:

@itemize @bullet
@item
To compile just the MySQL client libraries and client programs and
not the server, use the @code{--without-server} option:

@example
shell> ./configure --without-server
@end example

@c FIX incorrect..

If you don't have a C++ compiler, @code{mysql} will not compile (it is the
one client program that requires C++).  In this case,
you can remove the code in @code{configure} that tests for the C++ compiler
and then run @code{./configure} with the @code{--without-server} option. The
compile step will still try to build @code{mysql}, but you can ignore any
warnings about @file{mysql.cc}.  (If @code{make} stops, try @code{make -k}
to tell it to continue with the rest of the build even if errors occur.)

@item
If you want to get an embedded MySQL library (@code{libmysqld.a}) you should
use the @code{--with-embedded-server} option.

@item
If you don't want your log files and database directories located under
@file{/usr/local/var}, use a @code{configure} command, something like one
of these:

@example
shell> ./configure --prefix=/usr/local/mysql
shell> ./configure --prefix=/usr/local \
           --localstatedir=/usr/local/mysql/data
@end example

The first command changes the installation prefix so that everything is
installed under @file{/usr/local/mysql} rather than the default of
@file{/usr/local}.  The second command preserves the default installation
prefix, but overrides the default location for database directories
(normally @file{/usr/local/var}) and changes it to
@code{/usr/local/mysql/data}.  After you have compiled MySQL, you can
change these options with option files. @xref{Option files}.

@cindex changing socket location
@cindex socket location, changing
@cindex mysql.sock, changing location of
@item
If you are using Unix and you want the MySQL socket located somewhere
other than the default location (normally in the directory @file{/tmp} or
@file{/var/run}) use a @code{configure} command like this:

@example
shell> ./configure --with-unix-socket-path=/usr/local/mysql/tmp/mysql.sock
@end example

Note that the given file must be an absolute pathname!
You can also later change the location @file{mysql.sock} by using the MySQL
option files. @xref{Problems with mysql.sock}.

@cindex compiling, statically
@cindex statically, compiling
@item
If you want to compile statically linked programs (for example, to make a
binary distribution, to get more speed, or to work around problems with some
RedHat Linux distributions), run @code{configure} like this:

@example
shell> ./configure --with-client-ldflags=-all-static \
           --with-mysqld-ldflags=-all-static
@end example



@tindex @code{CC} environment variable
@tindex environment variable, @code{CC}
@tindex @code{CXX} environment variable
@tindex environment variable, @code{CXX}
@cindex @code{gcc}
@cindex C++ compiler, @code{gcc}
@cindex compiler, C++ @code{gcc}
@item
If you are using @code{gcc} and don't have @code{libg++} or @code{libstdc++}
installed, you can tell @code{configure} to use @code{gcc} as your C++
compiler:

@example
shell> CC=gcc CXX=gcc ./configure
@end example

When you use @code{gcc} as your C++ compiler, it will not attempt to link in
@code{libg++} or @code{libstdc++}.  This may be a good idea to do even if you
have the above libraries installed, as some versions of these libraries have
caused strange problems for MySQL users in the past.

Here are some common environment variables to set depending on
the compiler you are using:

@tindex CXXFLAGS environment variable
@tindex environment variable, CXXFLAGS
@multitable @columnfractions .19 .81
@item @strong{Compiler} @tab @strong{Recommended options}
@item gcc 2.7.2.1 @tab
CC=gcc CXX=gcc CXXFLAGS="-O3 -felide-constructors"
@item egcs 1.0.3a @tab
CC=gcc CXX=gcc CXXFLAGS="-O3 -felide-constructors -fno-exceptions -fno-rtti"
@item gcc 2.95.2 @tab
CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \
-felide-constructors -fno-exceptions -fno-rtti"
@item pgcc 2.90.29 or newer @tab
CFLAGS="-O3 -mpentiumpro -mstack-align-double" CXX=gcc \
CXXFLAGS="-O3 -mpentiumpro -mstack-align-double -felide-constructors \
-fno-exceptions -fno-rtti"
@end multitable

In most cases you can get a reasonably optimal MySQL binary by
using the options from the preceding table and adding the following options to the
configure line:

@example
--prefix=/usr/local/mysql --enable-assembler \
--with-mysqld-ldflags=-all-static
@end example

The full configure line would, in other words, be something like the
following for all recent gcc versions:

@example
CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \
-felide-constructors -fno-exceptions -fno-rtti" ./configure \
--prefix=/usr/local/mysql --enable-assembler \
--with-mysqld-ldflags=-all-static
@end example

The binaries we provide on the MySQL web site at
@uref{http://www.mysql.com/} are all compiled with full optimisation and
should be perfect for most users.  @xref{MySQL binaries}.  There are some
things you can tweak to make an even faster binary, but this is only for
advanced users.  @xref{Compile and link options}.

If the build fails and produces errors about your compiler or linker not
being able to create the shared library @file{libmysqlclient.so.#} (@samp{#}
is a version number), you can work around this problem by giving the
@code{--disable-shared} option to @code{configure}.  In this case,
@code{configure} will not build a shared @code{libmysqlclient.so.#} library.

@cindex default values, suppression
@cindex suppression, default values
@item
You can configure MySQL not to use @code{DEFAULT} column values for
non-@code{NULL} columns (that is, columns that are not allowed to be
@code{NULL}).  This causes @code{INSERT} statements to generate an error
unless you explicitly specify values for all columns that require a
non-@code{NULL} value.  To suppress use of default values, run
@code{configure} like this:

@tindex CXXFLAGS environment variable
@tindex environment variable, CXXFLAGS
@example
shell> CXXFLAGS=-DDONT_USE_DEFAULT_FIELDS ./configure
@end example

@cindex character sets
@findex configure option, --with-charset
@findex configure option, --with-extra-charset
@item
By default, MySQL uses the ISO-8859-1 (Latin1) character set. To
change the default set, use the @code{--with-charset} option:
@example
shell> ./configure --with-charset=CHARSET
@end example
@code{CHARSET} may be one of @code{big5}, @code{cp1251}, @code{cp1257},
@code{czech}, @code{danish}, @code{dec8}, @code{dos}, @code{euc_kr},
@code{gb2312}, @code{gbk}, @code{german1}, @code{hebrew}, @code{hp8},
@code{hungarian}, @code{koi8_ru}, @code{koi8_ukr}, @code{latin1},
@code{latin2}, @code{sjis}, @code{swe7}, @code{tis620}, @code{ujis},
@code{usa7}, or @code{win1251ukr}.
@xref{Character sets}.

If you want to convert characters between the server and the client,
you should take a look at the @code{SET OPTION CHARACTER SET} command.
@xref{SET OPTION, , @code{SET OPTION}}.

@cindex @code{myisamchk}
@strong{Warning}: If you change character sets after having created any
tables, you will have to run @code{myisamchk -r -q} on every table. Your
indexes may be sorted incorrectly otherwise.  (This can happen if you
install MySQL, create some tables, then reconfigure
MySQL to use a different character set and reinstall it.)

With the option @code{--with-extra-charset=LIST} you can define
which additional character sets should be compiled into the server.

Here @code{LIST} is either a list of character
sets separated with spaces,
@code{complex} to include all characters that can't be dynamically loaded,
or @code{all} to include all character sets into the binaries.

@item
To configure MySQL with debugging code, use the @code{--with-debug}
option:
@example
shell> ./configure --with-debug
@end example
This causes a safe memory allocator to be included that can find some errors
and that provides output about what is happening.
@xref{Debugging server}.

@item
If your client programs are using threads, you need to also compile a
thread-safe version of the MySQL client library with the
@code{--enable-thread-safe-client} configure options. This will create a
@code{libmysqlclient_r} library with which you should link your threaded
applications.  @xref{Threaded clients}.

@item
Options that pertain to particular systems can be found in the
system-specific section of this manual.
@xref{Operating System Specific Notes}.
@end itemize


@node Installing source tree, Compilation problems, configure options, Installing source
@subsection Installing from the Development Source Tree

@cindex development source tree
@cindex BitKeeper tree
@cindex cvs tree

@strong{Caution}: You should read this section only if you are interested
in helping us test our new code. If you just want to get MySQL up
and running on your system, you should use a standard release distribution
(either a source or binary distribution will do).

To obtain our most recent development source tree, use these instructions:

@enumerate
@item
Download @code{BitKeeper} from
@uref{http://www.bitmover.com/cgi-bin/download.cgi}.  You will need
@code{Bitkeeper} 2.0 or newer to access our repository.

@item
Follow the instructions to install it.

@item
After @code{BitKeeper} is installed, first go to the directory you
want to work from, and then use this command if you want to clone
the MySQL 3.23 branch:

@example
shell> bk clone bk://work.mysql.com:7000 mysql
@end example

To clone the 4.0 branch, use this command instead:

@example
shell> bk clone bk://work.mysql.com:7001 mysql-4.0
@end example

In the preceding examples the source tree will be set up in the @file{mysql/}
or @file{mysql-4.0/} subdirectory of your current directory.

The initial download of the source tree may take a while, depending on the
speed of your connection; be patient.

@item
You will need GNU @code{autoconf 2.52}, @code{automake 1.4},
@code{libtool}, and @code{m4} to run the next set of commands.

@code{automake} (1.5) doesn't yet work.

If you get some strange error during this stage, check that you really
have @code{libtool} installed!

@example
shell> cd mysql
shell> bk -r edit
shell> aclocal; autoheader; autoconf;  automake;
shell> ./configure  # Add your favorite options here
shell> make
@end example

A collection of our standard configure scripts is located in the
@file{BUILD/} subdirectory.  If you are lazy, you can use
@file{BUILD/compile-pentium-debug}. To compile on a different architecture,
modify the script by removing flags that are Pentium-specific.

@item
When the build is done, run @code{make install}.  Be careful with this
on a production machine; the command may overwrite your live release
installation.  If you have another installation of MySQL, we
recommand that you run @code{./configure} with different values for the
@code{prefix}, @code{with-tcp-port}, and @code{unix-socket-path} options than
those used for your production server.

@item
Play hard with your new installation and try to make the new features
crash.  Start by running @code{make test}.  @xref{MySQL test suite}.

@item
If you have gotten to the @code{make} stage and the distribution does
not compile, please report it to @email{bugs@@lists.mysql.com}.  If you
have installed the latest versions of the required GNU tools, and they
crash trying to process our configuration files, please report that also.
However, if you execute @code{aclocal} and get a @code{command not found}
error or a similar problem, do not report it.  Instead, make sure all
the necessary tools are installed and that your @code{PATH} variable is
set correctly so that your shell can find them.

@item
After the initial @code{bk clone} operation to get the source tree, you
should run @code{bk pull} periodically to get the updates.

@item
You can examine the change history for the tree with all the diffs by using
@code{bk sccstool}.  If you see some funny diffs or code that you have a
question about, do not hesitate to send e-mail to
@email{internals@@lists.mysql.com}.  Also, if you think you have a better idea
on how to do something, send an e-mail to the same address with a patch.
@code{bk diffs} will produce a patch for you after you have made changes
to the source. If you do not have the time to code your idea, just send
a description.

@item
@code{BitKeeper} has a nice help utility that you can access via
@code{bk helptool}.

@end enumerate


@node Compilation problems, MIT-pthreads, Installing source tree, Installing source
@subsection Problems Compiling?

@cindex compiling, problems
@cindex problems, compiling
@cindex reconfiguring
@cindex @code{config.cache} file
@cindex files, @code{config.cache}

All MySQL programs compile cleanly for us with no warnings on
Solaris using @code{gcc}.  On other systems, warnings may occur due to
differences in system include files.  See @ref{MIT-pthreads} for warnings
that may occur when using MIT-pthreads.  For other problems, check
the following list.

The solution to many problems involves reconfiguring.  If you do need to
reconfigure, take note of the following:

@cindex running @code{configure} after prior invocation
@cindex @code{configure}, running after prior invocation
@cindex reconfiguring
@cindex config.cache
@itemize @bullet
@item
If @code{configure} is run after it already has been run, it may use
information that was gathered during its previous invocation.  This
information is stored in @file{config.cache}.  When @code{configure} starts
up, it looks for that file and reads its contents if it exists, on the
assumption that the information is still correct.  That assumption is invalid
when you reconfigure.

@item
Each time you run @code{configure}, you must run @code{make} again
to recompile.  However, you may want to remove old object files from previous
builds first because they were compiled using different configuration options.
@end itemize

To prevent old configuration information or object files from being used,
run these commands before rerunning @code{configure}:

@example
shell> rm config.cache
shell> make clean
@end example

Alternatively, you can run @code{make distclean}.

The following list describes some of the problems when compiling MySQL
that have been found to occur most often:

@itemize @bullet
@item
@cindex @code{cc1plus} problems
@cindex @code{fatal signal 11}
@cindex @code{sql_yacc.cc} problems
@cindex internal compiler errors
@cindex virtual memory, problems while compiling
@cindex configure option, --with-low-memory
If you get errors when compiling @file{sql_yacc.cc}, such as the ones shown
here, you have probably run out of memory or swap space:

@example
Internal compiler error: program cc1plus got fatal signal 11
  or
Out of virtual memory
  or
Virtual memory exhausted
@end example

The problem is that @code{gcc} requires huge amounts of memory to compile
@file{sql_yacc.cc} with inline functions.  Try running @code{configure} with
the @code{--with-low-memory} option:

@example
shell> ./configure --with-low-memory
@end example

This option causes @code{-fno-inline} to be added to the compile line if you
are using @code{gcc} and @code{-O0} if you are using something else.  You
should try the @code{--with-low-memory} option even if you have so much
memory and swap space that you think you can't possibly have run out.  This
problem has been observed to occur even on systems with generous hardware
configurations, and the @code{--with-low-memory} option usually fixes it.

@item
By default, @code{configure} picks @code{c++} as the compiler name and
GNU @code{c++} links with @code{-lg++}.  If you are using @code{gcc},
that behavior can cause problems during configuration such as this:

@cindex C++ compiler cannot create executables
@example
configure: error: installation or configuration problem:
C++ compiler cannot create executables.
@end example

@tindex CXX environment variable
@tindex Environment variable, CXX
You might also observe problems during compilation related to
@code{g++}, @code{libg++}, or @code{libstdc++}.

One cause of these problems is that you may not have @code{g++}, or you may
have @code{g++} but not @code{libg++}, or @code{libstdc++}.  Take a look at
the @file{config.log} file.  It should contain the exact reason why your c++
compiler didn't work!  To work around these problems, you can use @code{gcc}
as your C++ compiler.  Try setting the environment variable @code{CXX} to
@code{"gcc -O3"}.  For example:

@tindex CXX environment variable
@tindex Environment variables, CXX
@example
shell> CXX="gcc -O3" ./configure
@end example

This works because @code{gcc} compiles C++ sources as well as @code{g++}
does, but does not link in @code{libg++} or @code{libstdc++} by default.

Another way to fix these problems, of course, is to install @code{g++},
@code{libg++}, and @code{libstdc++}.  We would however like to recommend
you to not use @code{libg++} or @code{libstdc++} with MySQL as this will
only increase the binary size of mysqld without giving you any benefits.
Some versions of these libraries have also caused strange problems for
MySQL users in the past.

@item
If your compile fails with errors, such as any of the following,
you must upgrade your version of @code{make} to GNU @code{make}:

@example
making all in mit-pthreads
make: Fatal error in reader: Makefile, line 18:
Badly formed macro assignment
  or
make: file `Makefile' line 18: Must be a separator (:
  or
pthread.h: No such file or directory
@end example

@cindex Solaris troubleshooting
@cindex FreeBSD troubleshooting
@cindex troubleshooting, Solaris
@cindex troubleshooting, FreeBSD
Solaris and FreeBSD are known to have troublesome @code{make} programs.

GNU @code{make} Version 3.75 is known to work.

@tindex CC environment variable
@tindex environment variable, CC
@tindex CXX environment variable
@tindex environment variable, CXX
@tindex CFLAGS environment variable
@tindex environment variable, CFLAGS
@tindex CXXFLAGS environment variable
@tindex environment variable, CXXFLAGS
@item
If you want to define flags to be used by your C or C++ compilers, do so by
adding the flags to the @code{CFLAGS} and @code{CXXFLAGS} environment
variables.  You can also specify the compiler names this way using @code{CC}
and @code{CXX}.  For example:

@example
shell> CC=gcc
shell> CFLAGS=-O3
shell> CXX=gcc
shell> CXXFLAGS=-O3
shell> export CC CFLAGS CXX CXXFLAGS
@end example

See @ref{MySQL binaries}, for a list of flag definitions that have been found
to be useful on various systems.

@item
If you get an error message like this,
you need to upgrade your @code{gcc} compiler:

@example
client/libmysql.c:273: parse error before `__attribute__'
@end example

@code{gcc} 2.8.1 is known to work, but we recommend using @code{gcc} 2.95.2 or
@code{egcs} 1.0.3a instead.

@item
If you get errors such as those shown here when compiling @code{mysqld},
@code{configure} didn't correctly detect the type of the last argument to
@code{accept()}, @code{getsockname()}, or @code{getpeername()}:

@example
cxx: Error: mysqld.cc, line 645: In this statement, the referenced
     type of the pointer value "&length" is "unsigned long", which
     is not compatible with "int".
new_sock = accept(sock, (struct sockaddr *)&cAddr, &length);
@end example

To fix this, edit the @file{config.h} file (which is generated by
@code{configure}).  Look for these lines:

@example
/* Define as the base type of the last arg to accept */
#define SOCKET_SIZE_TYPE XXX
@end example

Change @code{XXX} to @code{size_t} or @code{int}, depending on your
operating system.  (Note that you will have to do this each time you run
@code{configure} because @code{configure} regenerates @file{config.h}.)

@item
The @file{sql_yacc.cc} file is generated from @file{sql_yacc.yy}.  Normally
the build process doesn't need to create @file{sql_yacc.cc}, because
MySQL comes with an already generated copy.  However, if you do need
to re-create it, you might encounter this error:

@example
"sql_yacc.yy", line xxx fatal: default action causes potential...
@end example

This is a sign that your version of @code{yacc} is deficient.
You probably need to install @code{bison} (the GNU version of @code{yacc})
and use that instead.

@item
If you need to debug @code{mysqld} or a MySQL client, run
@code{configure} with the @code{--with-debug} option, then recompile and
link your clients with the new client library.  @xref{Debugging client}.
@end itemize


@node MIT-pthreads, Windows source build, Compilation problems, Installing source
@subsection MIT-pthreads Notes

@cindex MIT-pthreads
@cindex thread support, non-native

This section describes some of the issues involved in using MIT-pthreads.

Note that on Linux you should @strong{not} use MIT-pthreads but install LinuxThreads!
@xref{Linux}.

If your system does not provide native thread support, you will need to
build MySQL using the MIT-pthreads package.  This includes
older FreeBSD systems, SunOS 4.x, Solaris 2.4 and earlier, and some others.
@xref{Which OS}.

@itemize @bullet
@item
On most systems, you can force MIT-pthreads to be used by running
@code{configure} with the @code{--with-mit-threads} option:

@example
shell> ./configure --with-mit-threads
@end example

Building in a non-source directory is not supported when using
MIT-pthreads because we want to minimise our changes to this code.

@item
The checks that determine whether to use MIT-pthreads occur only
during the part of the configuration process that deals with the server
code.  If you have configured the distribution using @code{--without-server}
to build only the client code, clients will not know whether
MIT-pthreads is being used and will use Unix socket connections by default.
Because Unix sockets do not work under MIT-pthreads, this means you will need
to use @code{-h} or @code{--host} when you run client programs.

@item
When MySQL is compiled using MIT-pthreads, system locking is
disabled by default for performance reasons.  You can tell the server to use
system locking with the @code{--use-locking} option.

@item
Sometimes the pthread @code{bind()} command fails to bind to a socket without
any error message (at least on Solaris).  The result is that all connections
to the server fail.  For example:

@example
shell> mysqladmin version
mysqladmin: connect to server at '' failed;
error: 'Can't connect to mysql server on localhost (146)'
@end example

The solution to this is to kill the @code{mysqld} server and restart it.
This has only happened to us when we have forced the server down and done
a restart immediately.

@item
With MIT-pthreads, the @code{sleep()} system call isn't interruptible with
@code{SIGINT} (break).  This is only noticeable when you run
@code{mysqladmin --sleep}.  You must wait for the @code{sleep()} call to
terminate before the interrupt is served and the process stops.

@item
When linking, you may receive warning messages like these (at least on
Solaris); they can be ignored:

@example
ld: warning: symbol `_iob' has differing sizes:
    (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4;
file /usr/lib/libc.so value=0x140);
    /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken
ld: warning: symbol `__iob' has differing sizes:
    (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4;
file /usr/lib/libc.so value=0x140);
    /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken
@end example

@item
Some other warnings also can be ignored:

@example
implicit declaration of function `int strtoll(...)'
implicit declaration of function `int strtoul(...)'
@end example

@item
We haven't gotten @code{readline} to work with MIT-pthreads.  (This isn't
needed, but may be interesting for someone.)
@end itemize


@node Windows source build,  , MIT-pthreads, Installing source
@subsection Windows Source Distribution

You will need the following:

@itemize @bullet
@item
VC++ 6.0 compiler (updated with 4 or 5 SP and Pre-processor package)
The Pre-processor package is necessary for the macro assembler.
More details at:
@uref{http://msdn.microsoft.com/vstudio/sp/vs6sp5/faq.asp}.
@item
The MySQL source distribution for Windows, which can be downloaded
from @uref{http://www.mysql.com/downloads/}.
@end itemize

Building MySQL

@enumerate
@item
Create a work directory (e.g., workdir).
@item
Unpack the source distribution in the aforementioned directory.
@item
Start the VC++ 6.0 compiler.
@item
In the @code{File} menu, select @code{Open Workspace}.
@item
Open the @file{mysql.dsw} workspace you find on the work directory.
@item
From the @code{Build} menu,
select the @code{Set Active Configuration} menu.
@item
Click over the screen selecting @code{mysqld - Win32 Debug}
and click OK.
@item
Press @code{F7} to begin the build of the debug server, libs, and
some client applications.
@item
When the compilation finishes, copy the libs and the executables
to a separate directory.
@item
Compile the release versions that you want, in the same way.
@item
Create the directory for the MySQL stuff: e.g., @file{c:\mysql}
@item
From the workdir directory copy for the c:\mysql directory the
following directories:

@itemize @bullet
@item Data
@item Docs
@item Share
@end itemize

@item
Create the directory @file{c:\mysql\bin} and copy all the servers
and clients that you compiled previously.
@item
If you want, also create the @file{lib} directory and copy the
libs that you compiled previously.
@item
Do a clean using Visual Studio.
@end enumerate

Set up and start the server in the same way as for the binary
Windows distribution. @xref{Windows prepare environment}.


@node Post-installation, Upgrade, Installing source, Installing
@section Post-installation Setup and Testing

@cindex post-installation, setup and testing
@cindex testing, post-installation
@cindex setup, post-installation

@menu
* mysql_install_db::            Problems running @code{mysql_install_db}
* Starting server::             Problems starting the MySQL server
* Automatic start::             Starting and stopping MySQL automatically
@end menu

Once you've installed MySQL (from either a binary or source
distribution), you need to initialise the grant tables, start the server,
and make sure that the server works okay.  You may also wish to arrange
for the server to be started and stopped automatically when your system
starts up and shuts down.

Normally you install the grant tables and start the server like this
for installation from a source distribution:
@cindex starting, the server
@cindex server, starting

@example
shell> ./scripts/mysql_install_db
shell> cd mysql_installation_directory
shell> ./bin/safe_mysqld --user=mysql &
@end example

For a binary distribution (not RPM or pkg packages), do this:

@example
shell> cd mysql_installation_directory
shell> ./bin/mysql_install_db
shell> ./bin/safe_mysqld --user=mysql &
@end example

This creates the @code{mysql} database which will hold all database
privileges, the @code{test} database which you can use to test
MySQL, and also privilege entries for the user that run
@code{mysql_install_db} and a @code{root} user (without any passwords).
This also starts the @code{mysqld} server.

@code{mysql_install_db} will not overwrite any old privilege tables, so
it should be safe to run in any circumstances.  If you don't want to
have the @code{test} database you can remove it with @code{mysqladmin -u
root drop test}.

Testing is most easily done from the top-level directory of the MySQL
distribution.  For a binary distribution, this is your installation directory
(typically something like @file{/usr/local/mysql}).  For a source
distribution, this is the main directory of your MySQL source tree.
@cindex testing, the server

In the commands shown in this section and in the following
subsections, @code{BINDIR} is the path to the location in which programs
like @code{mysqladmin} and @code{safe_mysqld} are installed.  For a
binary distribution, this is the @file{bin} directory within the
distribution.  For a source distribution, @code{BINDIR} is probably
@file{/usr/local/bin}, unless you specified an installation directory
other than @file{/usr/local} when you ran @code{configure}.
@code{EXECDIR} is the location in which the @code{mysqld} server is
installed.  For a binary distribution, this is the same as
@code{BINDIR}.  For a source distribution, @code{EXECDIR} is probably
@file{/usr/local/libexec}.

Testing is described in detail:
@cindex testing, installation

@enumerate
@item
If necessary, start the @code{mysqld} server and set up the initial
MySQL grant tables containing the privileges that determine how
users are allowed to connect to the server.  This is normally done with the
@code{mysql_install_db} script:

@example
shell> scripts/mysql_install_db
@end example

Typically, @code{mysql_install_db} needs to be run only the first time you
install MySQL.  Therefore, if you are upgrading an existing
installation, you can skip this step.  (However, @code{mysql_install_db} is
quite safe to use and will not update any tables that already exist, so if
you are unsure of what to do, you can always run @code{mysql_install_db}.)

@code{mysql_install_db} creates six tables (@code{user}, @code{db},
@code{host}, @code{tables_priv}, @code{columns_priv}, and @code{func}) in the
@code{mysql} database.  A description of the initial privileges is given in
@ref{Default privileges}.  Briefly, these privileges allow the MySQL
@code{root} user to do anything, and allow anybody to create or use databases
with a name of @code{test} or starting with @code{test_}.

If you don't set up the grant tables, the following error will appear in the
log file when you start the server:

@tindex host.frm, problems finding
@example
mysqld: Can't find file: 'host.frm'
@end example

This may also happen with a binary MySQL distribution if you
don't start MySQL by executing exactly @code{./bin/safe_mysqld}!
@xref{safe_mysqld, , @code{safe_mysqld}}.

You might need to run @code{mysql_install_db} as @code{root}.  However,
if you prefer, you can run the MySQL server as an unprivileged
(non-@code{root}) user, provided that the user can read and write files in
the database directory.  Instructions for running MySQL as an
unprivileged user are given in @ref{Changing MySQL user, , Changing
MySQL user}.

If you have problems with @code{mysql_install_db}, see
@ref{mysql_install_db, , @code{mysql_install_db}}.

There are some alternatives to running the @code{mysql_install_db}
script as it is provided in the MySQL distribution:

@itemize @bullet
@item
You may want to edit @code{mysql_install_db} before running it, to change
the initial privileges that are installed into the grant tables.  This is
useful if you want to install MySQL on a lot of machines with the
same privileges.  In this case you probably should need only to add a few
extra @code{INSERT} statements to the @code{mysql.user} and @code{mysql.db}
tables!

@item
If you want to change things in the grant tables after installing them, you
can run @code{mysql_install_db}, then use @code{mysql -u root mysql} to
connect to the grant tables as the MySQL @code{root} user and issue
SQL statements to modify the grant tables directly.

@item
It is possible to re-create the grant tables completely after they have
already been created.  You might want to do this if you've already installed
the tables but then want to re-create them after editing
@code{mysql_install_db}.
@end itemize

For more information about these alternatives, see @ref{Default privileges}.

@item
Start the MySQL server like this:

@example
shell> cd mysql_installation_directory
shell> bin/safe_mysqld &
@end example

If you have problems starting the server, see @ref{Starting server}.

@item
Use @code{mysqladmin} to verify that the server is running.  The following
commands provide a simple test to check that the server is up and responding
to connections:

@example
shell> BINDIR/mysqladmin version
shell> BINDIR/mysqladmin variables
@end example

The output from @code{mysqladmin version} varies slightly depending on your
platform and version of MySQL, but should be similar to that shown here:

@example
shell> BINDIR/mysqladmin version
mysqladmin  Ver 8.14 Distrib 3.23.32, for linux on i586
Copyright (C) 2000 MySQL AB & MySQL Finland AB & TCX DataKonsult AB
This software comes with ABSOLUTELY NO WARRANTY. This is free software,
and you are welcome to modify and redistribute it under the GPL license.

Server version          3.23.32-debug
Protocol version        10
Connection              Localhost via Unix socket
TCP port                3306
UNIX socket             /tmp/mysql.sock
Uptime:                 16 sec

Threads: 1  Questions: 9  Slow queries: 0
Opens: 7  Flush tables: 2  Open tables: 0
Queries per second avg: 0.000
Memory in use: 132K  Max memory used: 16773K
@end example

To get a feeling for what else you can do with @code{BINDIR/mysqladmin},
invoke it with the @code{--help} option.

@item
Verify that you can shut down the server:
@cindex server, shutdown
@cindex shutting down, the server

@example
shell> BINDIR/mysqladmin -u root shutdown
@end example

@item
Verify that you can restart the server.  Do this using @code{safe_mysqld} or
by invoking @code{mysqld} directly.  For example:

@cindex server, restart
@cindex restarting, the server

@example
shell> BINDIR/safe_mysqld --log &
@end example

If @code{safe_mysqld} fails, try running it from the MySQL
installation directory (if you are not already there).  If that doesn't work,
see @ref{Starting server}.

@item
Run some simple tests to verify that the server is working.
The output should be similar to what is shown here:

@example
shell> BINDIR/mysqlshow
+-----------+
| Databases |
+-----------+
| mysql     |
+-----------+

shell> BINDIR/mysqlshow mysql
Database: mysql
+--------------+
|    Tables    |
+--------------+
| columns_priv |
| db           |
| func         |
| host         |
| tables_priv  |
| user         |
+--------------+

shell> BINDIR/mysql -e "SELECT host,db,user FROM db" mysql
+------+--------+------+
| host | db     | user |
+------+--------+------+
| %    | test   |      |
| %    | test_% |      |
+------+--------+------+
@end example

There is also a benchmark suite in the @file{sql-bench} directory (under the
MySQL installation directory) that you can use to compare how
MySQL performs on different platforms.  The @file{sql-bench/Results}
directory contains the results from many runs against different databases and
platforms.  To run all tests, execute these commands:

@example
shell> cd sql-bench
shell> run-all-tests
@end example

If you don't have the @file{sql-bench} directory, you are probably using an
RPM for a binary distribution.  (Source distribution RPMs include the
benchmark directory.)  In this case, you must first install the benchmark
suite before you can use it.  Beginning with MySQL Version 3.22,
there are benchmark RPM files named @file{mysql-bench-VERSION-i386.rpm} that
contain benchmark code and data.

If you have a source distribution, you can also run the tests in the
@file{tests} subdirectory. For example, to run @file{auto_increment.tst}, do
this:

@example
shell> BINDIR/mysql -vvf test < ./tests/auto_increment.tst
@end example

The expected results are shown in the @file{./tests/auto_increment.res} file.
@end enumerate


@node mysql_install_db, Starting server, Post-installation, Post-installation
@subsection Problems Running @code{mysql_install_db}
@cindex @code{mysql_install_db} script
@cindex scripts, @code{mysql_install_db}

The purpose of the @code{mysql_install_db} script is to generate new
MySQL privilege tables.  It will not affect any other data!
It will also not do anything if you already have MySQL privilege
tables installed!

If you want to re-create your privilege tables, you should take down
the @code{mysqld} server, if it's running, and then do something like:

@example
mv mysql-data-directory/mysql mysql-data-directory/mysql-old
mysql_install_db
@end example

This section lists problems you might encounter when you run
@code{mysql_install_db}:

@table @strong
@item @code{mysql_install_db} doesn't install the grant tables

You may find that @code{mysql_install_db} fails to install the grant
tables and terminates after displaying the following messages:

@example
starting mysqld daemon with databases from XXXXXX
mysql daemon ended
@end example

In this case, you should examine the log file very carefully!  The log
should be located in the directory @file{XXXXXX} named by the error message,
and should indicate why @code{mysqld} didn't start.  If you don't understand
what happened, include the log when you post a bug report using
@code{mysqlbug}!
@xref{Bug reports}.

@item There is already a @code{mysqld} daemon running

In this case, you probably don't have to run @code{mysql_install_db} at
all.  You have to run @code{mysql_install_db} only once, when you install
MySQL the first time.

@item Installing a second @code{mysqld} daemon doesn't work when one daemon is running

This can happen when you already have an existing MySQL
installation, but want to put a new installation in a different place (for
example, for testing, or perhaps you simply want to run two installations at
the same time).  Generally the problem that occurs when you try to run the
second server is that it tries to use the same socket and port as the old one.
In this case you will get the error message: @code{Can't start server: Bind on
TCP/IP port: Address already in use} or @code{Can't start server: Bind on
unix socket...}. @xref{Installing many servers}.

@item You don't have write access to @file{/tmp}
@cindex write access, tmp
@cindex temporary file, write access
@cindex files, @code{tmp}

If you don't have write access to create a socket file at the default place
(in @file{/tmp}) or permission to create temporary files in @file{/tmp,}
you will get an error when running @code{mysql_install_db} or when
starting or using @code{mysqld}.

You can specify a different socket and temporary directory as follows:

@tindex TMPDIR environment variable
@tindex MYSQL_UNIX_PORT environment variable
@tindex Environment variable, TMPDIR
@tindex Environment variable, MYSQL_UNIX_PORT
@example
shell> TMPDIR=/some_tmp_dir/
shell> MYSQL_UNIX_PORT=/some_tmp_dir/mysqld.sock
shell> export TMPDIR MYSQL_UNIX_PORT
@end example

See @ref{Problems with mysql.sock}.

@file{some_tmp_dir} should be the path to some directory for which you
have write permission. @xref{Environment variables}.

After this you should be able to run @code{mysql_install_db} and start
the server with these commands:

@example
shell> scripts/mysql_install_db
shell> BINDIR/safe_mysqld &
@end example

@item @code{mysqld} crashes immediately

If you are running RedHat Version 5.0 with a version of @code{glibc} older than
2.0.7-5, you should make sure you have installed all @code{glibc} patches!
There is a lot of information about this in the MySQL mail
archives.  Links to the mail archives are available online at
@uref{http://lists.mysql.com/}.
Also, see @ref{Linux}.

You can also start @code{mysqld} manually using the @code{--skip-grant-tables}
option and add the privilege information yourself using @code{mysql}:

@example
shell> BINDIR/safe_mysqld --skip-grant-tables &
shell> BINDIR/mysql -u root mysql
@end example

From @code{mysql}, manually execute the SQL commands in
@code{mysql_install_db}.  Make sure you run @code{mysqladmin
flush-privileges} or @code{mysqladmin reload} afterward to tell the server to
reload the grant tables.
@end table


@node Starting server, Automatic start, mysql_install_db, Post-installation
@subsection Problems Starting the MySQL Server
@cindex server, starting problems
@cindex problems, starting the server

If you are going to use tables that support transactions (InnoDB, BDB),
you should first create a my.cnf file and set startup options
for the table types you plan to use. @xref{Table types}.

Generally, you start the @code{mysqld} server in one of these ways:

@itemize @bullet
@item
By invoking @code{mysql.server}.  This script is used primarily at
system startup and shutdown, and is described more fully in
@ref{Automatic start}.

@item
By invoking @code{safe_mysqld}, which tries to determine the proper options
for @code{mysqld} and then runs it with those options. @xref{safe_mysqld, ,
@code{safe_mysqld}}.

@item
For Windows NT/2000/XP, please see @ref{NT start}.

@item
By invoking @code{mysqld} directly.
@end itemize

When the @code{mysqld} daemon starts up, it changes the directory to the
data directory.  This is where it expects to write log files and the pid
(process ID) file, and where it expects to find databases.

The data directory location is hardwired in when the distribution is
compiled.  However, if @code{mysqld} expects to find the data directory
somewhere other than where it really is on your system, it will not work
properly.  If you have problems with incorrect paths, you can find out
what options @code{mysqld} allows and what the default path settings are by
invoking @code{mysqld} with the @code{--help} option.  You can override the
defaults by specifying the correct pathnames as command-line arguments to
@code{mysqld}.  (These options can be used with @code{safe_mysqld} as well.)

Normally you should need to tell @code{mysqld} only the base directory under
which MySQL is installed.  You can do this with the @code{--basedir}
option.  You can also use @code{--help} to check the effect of changing path
options (note that @code{--help} @strong{must} be the final option of the
@code{mysqld} command).  For example:

@example
shell> EXECDIR/mysqld --basedir=/usr/local --help
@end example

Once you determine the path settings you want, start the server without
the @code{--help} option.

Whichever method you use to start the server, if it fails to start up
correctly, check the log file to see if you can find out why.  Log files
are located in the data directory (typically
@file{/usr/local/mysql/data} for a binary distribution,
@file{/usr/local/var} for a source distribution, and
@file{\mysql\data\mysql.err} on Windows).  Look in the data directory for
files with names of the form @file{host_name.err} and
@file{host_name.log} where @code{host_name} is the name of your server
host.  Then check the last few lines of these files:

@example
shell> tail host_name.err
shell> tail host_name.log
@end example

Look for something like the following in the log file:
@example
000729 14:50:10  bdb:  Recovery function for LSN 1 27595 failed
000729 14:50:10  bdb:  warning: ./test/t1.db: No such file or directory
000729 14:50:10  Can't init databases
@end example

This means that you didn't start @code{mysqld} with @code{--bdb-no-recover}
and Berkeley DB found something wrong with its log files when it
tried to recover your databases.  To be able to continue, you should
move away the old Berkeley DB log file from the database directory to
some other place, where you can later examine it.  The log files are
named @file{log.0000000001}, where the number will increase over time.

If you are running @code{mysqld} with BDB table support and @code{mysqld} core
dumps at start this could be because of some problems with the BDB
recover log.  In this case you can try starting @code{mysqld} with
@code{--bdb-no-recover}.  If this helps, then you should remove all
@file{log.*} files from the data directory and try starting @code{mysqld}
again.

If you get the following error, it means that some other program (or another
@code{mysqld} server) is already using the TCP/IP port or socket
@code{mysqld} is trying to use:

@example
Can't start server: Bind on TCP/IP port: Address already in use
  or
Can't start server : Bind on unix socket...
@end example

Use @code{ps} to make sure that you don't have another @code{mysqld} server
running.  If you can't find another server running, you can try to execute
the command @code{telnet your-host-name tcp-ip-port-number} and press
Enter a couple of times.  If you don't get an error message like
@code{telnet: Unable to connect to remote host: Connection refused},
something is using the TCP/IP port @code{mysqld} is trying to use.
See @ref{mysql_install_db} and @ref{Multiple servers}.

If @code{mysqld} is currently running, you can find out what path settings
it is using by executing this command:

@example
shell> mysqladmin variables
@end example

or

@example
shell> mysqladmin -h 'your-host-name' variables
@end example

If you get @code{Errcode 13}, which means @code{Permission denied}, when
starting @code{mysqld} this means that you didn't have the right to
read/create files in the MySQL database or log directory. In this case
you should either start @code{mysqld} as the root user or change the
permissions for the involved files and directories so that you have the
right to use them.

If @code{safe_mysqld} starts the server but you can't connect to it,
you should make sure you have an entry in @file{/etc/hosts} that looks like
this:

@example
127.0.0.1       localhost
@end example

This problem occurs only on systems that don't have a working thread
library and for which MySQL must be configured to use MIT-pthreads.

If you can't get @code{mysqld} to start you can try to make a trace file
to find the problem. @xref{Making trace files}.

If you are using InnoDB tables, refer to the InnoDB-specific startup
options.  @xref{InnoDB start}.

If you are using BDB (Berkeley DB) tables, you should familiarise
yourself with the different BDB specific startup options.  @xref{BDB start}.


@node Automatic start,  , Starting server, Post-installation
@subsection Starting and Stopping MySQL Automatically

@cindex starting, the server automatically
@cindex stopping, the server
@cindex server, starting and stopping

The @code{mysql.server} and @code{safe_mysqld} scripts can be used to start
the server automatically at system startup time. @code{mysql.server} can also
be used to stop the server.

The @code{mysql.server} script can be used to start or stop the server
by invoking it with @code{start} or @code{stop} arguments:

@example
shell> mysql.server start
shell> mysql.server stop
@end example

@code{mysql.server} can be found in the @file{share/mysql} directory
under the MySQL installation directory or in the @file{support-files}
directory of the MySQL source tree.

Before @code{mysql.server} starts the server, it changes the directory to
the MySQL installation directory, then invokes @code{safe_mysqld}.
You might need to edit @code{mysql.server} if you have a binary distribution
that you've installed in a non-standard location.  Modify it to @code{cd}
into the proper directory before it runs @code{safe_mysqld}. If you want the
server to run as some specific user, add an appropriate @code{user} line
to the @file{/etc/my.cnf} file, as shown later in this section.

@code{mysql.server stop} brings down the server by sending a signal to it.
You can take down the server manually by executing @code{mysqladmin shutdown}.

You might want to add these start and stop commands to the appropriate places
in your @file{/etc/rc*} files when you start using MySQL for
production applications.  Note that if you modify @code{mysql.server}, and then
upgrade MySQL sometime, your modified version will be overwritten,
so you should make a copy of your edited version that you can reinstall.

If your system uses @file{/etc/rc.local} to start external scripts, you
should append the following to it:
@example
/bin/sh -c 'cd /usr/local/mysql ; ./bin/safe_mysqld --user=mysql &'
@end example

@cindex changing socket location
You can also add options for @code{mysql.server} in a global
@file{/etc/my.cnf} file.  A typical @file{/etc/my.cnf} file might look like
this:

@example
[mysqld]
datadir=/usr/local/mysql/var
socket=/var/tmp/mysql.sock
port=3306
user=mysql

[mysql.server]
basedir=/usr/local/mysql
@end example

The @code{mysql.server} script understands the following options:
@code{datadir}, @code{basedir}, and @code{pid-file}.

The following table shows which option groups each of the startup scripts
read from option files:

@multitable @columnfractions .15 .35
@item @strong{Script} @tab @strong{Option groups}
@item @code{mysqld} @tab @code{mysqld} and @code{server}
@item @code{mysql.server} @tab @code{mysql.server}, @code{mysqld}, and @code{server}
@item @code{safe_mysqld} @tab @code{mysql.server}, @code{mysqld}, and @code{server}
@end multitable

@xref{Option files}.


@node Upgrade, Operating System Specific Notes, Post-installation, Installing
@section Upgrading/Downgrading MySQL

@cindex upgrading
@cindex downgrading

You can always move the MySQL form and datafiles between
different versions on the same architecture as long as you have the same
base version of MySQL. The current base version is
3. If you change the character set when running MySQL (which may
also change the sort order), you must run @code{myisamchk -r -q} on all
tables.  Otherwise, your indexes may not be ordered correctly.

If you are afraid of new versions, you can always rename your old
@code{mysqld} to something like @code{mysqld-old-version-number}.  If
your new @code{mysqld} then does something unexpected, you can simply shut it
down and restart with your old @code{mysqld}!

When you do an upgrade you should also back up your old databases, of course.

If after an upgrade, you experience problems with recompiled client programs,
like @code{Commands out of sync} or unexpected core dumps, you probably have
used an old header or library file when compiling your programs.  In this
case you should check the date for your @file{mysql.h} file and
@file{libmysqlclient.a} library to verify that they are from the new
MySQL distribution.  If not, please recompile your programs!

If you get some problems that the new @code{mysqld} server doesn't want to
start or that you can't connect without a password, check that you don't
have some old @file{my.cnf} file from your old installation!  You can
check this with: @code{program-name --print-defaults}.  If this outputs
anything other than the program name, you have an active @file{my.cnf}
file that will affect things!

It is a good idea to rebuild and reinstall the @code{Msql-Mysql-modules}
distribution whenever you install a new release of MySQL,
particularly if you notice symptoms such as all your @code{DBI} scripts
dumping core after you upgrade MySQL.

@menu
* Upgrading-from-3.23::         Upgrading from a 3.23 version to 4.0
* Upgrading-from-3.22::         Upgrading from a 3.22 version to 3.23
* Upgrading-from-3.21::         Upgrading from a 3.21 version to 3.22
* Upgrading-from-3.20::         Upgrading from a 3.20 version to 3.21
* Upgrading-to-arch::           Upgrading to another architecture
@end menu

@cindex compatibility, between MySQL versions
@cindex upgrading, 3.23 to 4.0
@node Upgrading-from-3.23, Upgrading-from-3.22, Upgrade, Upgrade
@subsection Upgrading From Version 3.23 to Version 4.0

You can use your old datafiles without any modification with Version 4.0.
If you want to move your data from a MySQL 4.0 server to an older server,
you have to use @code{mysqldump}.

Old clients should work with a Version 4.0 server without any problems.

The following lists tell what you have to watch out for when upgrading to
version 4.0;

@itemize @bullet
@item
@code{DOUBLE} and @code{FLOAT} columns are now honoring the
@code{UNSIGNED} flag on storage (before, @code{UNSIGNED} was ignored for
these columns).
@item
@code{ORDER BY column DESC} now always sorts @code{NULL} values
first; in 3.23 this was not always consistent.
@item
@code{SHOW INDEX} has 2 columns more (@code{Null} and @code{Index_type})
than it had in 3.23.
@item
@code{SIGNED} is a reserved word.
@item
The result of all bitwise operators @code{|}, @code{&}, @code{<<},
@code{>>}, and @code{~} is now unsigned.  This may cause problems if you
are using them in a context where you want a signed result.
@xref{Cast Functions}.
@item
@strong{Note}: when you use subtraction between integer values where
one is of type @code{UNSIGNED}, the result will be unsigned!  In other
words, before upgrading to MySQL 4.0, you should check your application
for cases where you are subtracting a value from an unsigned entity and
want a negative answer or subtracting an unsigned value from an
integer column. You can disable this behaviour by using the
@code{--sql-mode=NO_UNSIGNED_SUBTRACTION} option when starting
@code{mysqld}.  @xref{Cast Functions}.
@item
To use @code{MATCH ... AGAINST (... IN BOOLEAN MODE)} with your tables,
you need to rebuild them with @code{ALTER TABLE table_name TYPE=MyISAM},
@strong{even} if they are of @code{MyISAM} type.
@item
@code{LOCATE()} and @code{INSTR()} are case-sensitive if one of the
arguments is a binary string.
@item
@code{STRCMP()} now uses the current character set when doing comparisons,
which means that the default comparison behavior now is case-insensitive.
@item
@code{HEX(string)} now returns the characters in string converted to
hexadecimal.  If you want to convert a number to hexadecimal, you should
ensure that you call @code{HEX()} with a numeric argument.
@item
In 3.23, @code{INSERT INTO ... SELECT} always had @code{IGNORE} enabled.
In 4.0.1, MySQL will stop (and possibly roll back) in case of an error if you
don't specify @code{IGNORE}.
@item
@file{safe_mysqld} is renamed to @file{mysqld_safe}.
@item
The old C API functions @code{mysql_drop_db}, @code{mysql_create_db}, and
@code{mysql_connect} are not supported anymore, unless you compile
MySQL with @code{CFLAGS=-DUSE_OLD_FUNCTIONS}.  Instead of doing this,
it is preferable to change the client to use the new 4.0 API.
@item
In the @code{MYSQL_FIELD} structure, @code{length} and @code{max_length} have
changed from @code{unsigned int} to @code{unsigned long}. This should not
cause any problems, except that they may generate warning messages when
used as arguments in the @code{printf()} class of functions.
@item
You should use @code{TRUNCATE TABLE} when you want to delete all rows
from a table and you don't care how many rows were deleted.
(Because @code{TRUNCATE TABLE} is faster than @code{DELETE FROM table_name}).
@item
You will get an error if you have an active @code{LOCK TABLES} or
transaction when trying to execute @code{TRUNCATE TABLE} or @code{DROP
DATABASE}.
@item
You should use integers to store values in BIGINT columns (instead of using
strings, as you did in MySQL 3.23).  Using strings will still work, but using
integers is more efficient.
@item
Format of @code{SHOW OPEN TABLE} has changed.
@item
Multi-threaded clients should use @code{mysql_thread_init()} and
@code{mysql_thread_end()}. @xref{Threaded clients}.
@item
If you want to recompile the Perl DBD::mysql module, you must get
Msql-Mysql-modules version 1.2218 or newer because the older DBD modules
used the deprecated @code{drop_db()} call.
@item
@code{RAND(seed)} returns a different random number series in 4.0 than in
3.23; this was done to further differentiate @code{RAND(seed)} and @code{RAND(seed+1)}.
@end itemize

@node Upgrading-from-3.22, Upgrading-from-3.21, Upgrading-from-3.23, Upgrade
@subsection Upgrading From Version 3.22 to Version 3.23

@cindex compatibility, between MySQL versions
@cindex upgrading, 3.22 to 3.23

MySQL Version 3.23 supports tables of the new @code{MyISAM} type and
the old @code{ISAM} type.  You don't have to convert your old tables to
use these with Version 3.23.  By default, all new tables will be created with
type @code{MyISAM} (unless you start @code{mysqld} with the
@code{--default-table-type=isam} option). You can change an @code{ISAM}
table to a @code{MyISAM} table with @code{ALTER TABLE table_name TYPE=MyISAM}
or the Perl script @code{mysql_convert_table_format}.

Version 3.22 and 3.21 clients will work without any problems with a Version
3.23 server.

The following list tells what you have to watch out for when upgrading to
Version 3.23:

@itemize @bullet
@item
All tables that use the @code{tis620} character set must be fixed
with @code{myisamchk -r} or @code{REPAIR TABLE}.

@item
If you do a @code{DROP DATABASE} on a symbolic linked database, both the
link and the original database are deleted.  (This didn't happen in 3.22
because configure didn't detect the @code{readlink} system call.)

@item
@code{OPTIMIZE TABLE} now only works for @code{MyISAM} tables.
For other table types, you can use @code{ALTER TABLE} to optimise the table.
During @code{OPTIMIZE TABLE} the table is now locked from other threads.

@item
The MySQL client @code{mysql} is now by default started with the
option @code{--no-named-commands (-g)}. This option can be disabled with
@code{--enable-named-commands (-G)}. This may cause incompatibility problems in
some cases@-for example, in SQL scripts that use named commands without a
semicolon!  Long format commands still work from the first line.

@item
Date functions that work on parts of dates (like @code{MONTH()}) will now
return 0 for @code{0000-00-00} dates. (MySQL 3.22 returned @code{NULL}.)

@item
If you are using the @code{german} character sort order, you must repair
all your tables with @code{isamchk -r}, as we have made some changes in
the sort order!

@item
The default return type of @code{IF} will now depend on both arguments
and not only the first argument.

@item
@code{AUTO_INCREMENT} will not work with negative numbers. The reason
for this is that negative numbers caused problems when wrapping from -1 to 0.
@code{AUTO_INCREMENT} for MyISAM tables is no handled at a lower level and
is much faster than before. For MyISAM tables old numbers are also not reused
anymore, even if you delete some rows from the table.

@item
@code{CASE}, @code{DELAYED}, @code{ELSE}, @code{END}, @code{FULLTEXT},
@code{INNER}, @code{RIGHT}, @code{THEN}, and @code{WHEN} are now reserved words.

@item
@code{FLOAT(X)} is now a true floating-point type and not a value with a
fixed number of decimals.

@item
When declaring @code{DECIMAL(length,dec)} the length argument no longer
includes a place for the sign or the decimal point.

@item
A @code{TIME} string must now be of one of the following formats:
@code{[[[DAYS] [H]H:]MM:]SS[.fraction]} or
@code{[[[[[H]H]H]H]MM]SS[.fraction]}.

@item
@code{LIKE} now compares strings using the same character comparison rules
as @code{=}.  If you require the old behavior, you can compile
MySQL with the @code{CXXFLAGS=-DLIKE_CMP_TOUPPER} flag.

@item
@code{REGEXP} is now case-insensitive for normal (not binary) strings.

@item
When you check/repair tables you should use @code{CHECK TABLE}
or @code{myisamchk} for @code{MyISAM} tables (@file{.MYI}) and
@code{isamchk} for ISAM (@file{.ISM}) tables.

@item
If you want your @code{mysqldump} files to be compatible between
MySQL Version 3.22 and Version 3.23, you should not use the
@code{--opt} or @code{--full} option to @code{mysqldump}.

@item
Check all your calls to @code{DATE_FORMAT()} to make sure there is a
@samp{%} before each format character.
(MySQL Version 3.22 and later already allowed this syntax.)

@item
@code{mysql_fetch_fields_direct} is now a function (it was a macro) and
it returns a pointer to a @code{MYSQL_FIELD} instead of a
@code{MYSQL_FIELD}.

@item
@code{mysql_num_fields()} can no longer be used on a @code{MYSQL*} object (it's
now a function that takes @code{MYSQL_RES*} as an argument, so you should
use @code{mysql_field_count()} instead).

@item
In MySQL Version 3.22, the output of @code{SELECT DISTINCT ...} was
almost always sorted.  In Version 3.23, you must use @code{GROUP BY} or
@code{ORDER BY} to obtain sorted output.

@item
@code{SUM()} now returns @code{NULL}, instead of 0, if
there are no matching rows. This is according to ANSI SQL.

@item
An @code{AND} or @code{OR} with @code{NULL} values will now return
@code{NULL} instead of 0. This mostly affects queries that use @code{NOT}
on an @code{AND/OR} expression as @code{NOT NULL} = @code{NULL}.
@code{LPAD()} and @code{RPAD()} will shorten the result string if it's longer
than the length argument.
@end itemize


@node Upgrading-from-3.21, Upgrading-from-3.20, Upgrading-from-3.22, Upgrade
@subsection Upgrading from Version 3.21 to Version 3.22

@cindex compatibility, between MySQL versions
@cindex upgrading, 3.21 to 3.22

Nothing that affects compatibility has changed between versions 3.21 and 3.22.
The only pitfall is that new tables that are created with @code{DATE} type
columns will use the new way to store the date. You can't access these new
fields from an old version of @code{mysqld}.

After installing MySQL Version 3.22, you should start the new server
and then run the @code{mysql_fix_privilege_tables} script. This will add the
new privileges that you need to use the @code{GRANT} command.  If you forget
this, you will get @code{Access denied} when you try to use @code{ALTER
TABLE}, @code{CREATE INDEX}, or @code{DROP INDEX}. If your MySQL root
user requires a password, you should give this as an argument to
@code{mysql_fix_privilege_tables}.

The C API interface to @code{mysql_real_connect()} has changed.  If you have
an old client program that calls this function, you must place a @code{0} for
the new @code{db} argument (or recode the client to send the @code{db}
element for faster connections).  You must also call @code{mysql_init()}
before calling @code{mysql_real_connect()}!  This change was done to allow
the new @code{mysql_options()} function to save options in the @code{MYSQL}
handler structure.

The @code{mysqld} variable @code{key_buffer} has changed names to
@code{key_buffer_size}, but you can still use the old name in your
startup files.


@node Upgrading-from-3.20, Upgrading-to-arch, Upgrading-from-3.21, Upgrade
@subsection Upgrading from Version 3.20 to Version 3.21

@cindex upgrading, 3.20 to 3.21

If you are running a version older than Version 3.20.28 and want to
switch to Version 3.21, you need to do the following:

You can start the @code{mysqld} Version 3.21 server with @code{safe_mysqld
--old-protocol} to use it with clients from a Version 3.20 distribution.
In this case, the new client function @code{mysql_errno()} will not
return any server error, only @code{CR_UNKNOWN_ERROR} (but it
works for client errors), and the server uses the old @code{password()}
checking rather than the new one.

If you are @strong{not} using the @code{--old-protocol} option to
@code{mysqld}, you will need to make the following changes:

@itemize @bullet
@item
All client code must be recompiled. If you are using ODBC, you must get
the new @code{MyODBC} 2.x driver.
@item
The script @code{scripts/add_long_password} must be run to convert the
@code{Password} field in the @code{mysql.user} table to @code{CHAR(16)}.
@item
All passwords must be reassigned in the @code{mysql.user} table (to get 62-bit
rather than 31-bit passwords).
@item
The table format hasn't changed, so you don't have to convert any tables.
@end itemize

MySQL Version 3.20.28 and above can handle the new @code{user} table
format without affecting clients. If you have a MySQL version earlier
than Version 3.20.28, passwords will no longer work with it if you convert the
@code{user} table. So to be safe, you should first upgrade to at least Version
3.20.28 and then upgrade to Version 3.21.

@cindex Protocol mismatch
The new client code works with a 3.20.x @code{mysqld} server, so
if you experience problems with 3.21.x, you can use the old 3.20.x server
without having to recompile the clients again.

If you are not using the @code{--old-protocol} option to @code{mysqld},
old clients will issue the error message:

@example
ERROR: Protocol mismatch. Server Version = 10 Client Version = 9
@end example

The new Perl @code{DBI}/@code{DBD} interface also supports the old
@code{mysqlperl} interface.  The only change you have to make if you use
@code{mysqlperl} is to change the arguments to the @code{connect()} function.
The new arguments are: @code{host}, @code{database}, @code{user},
and @code{password} (the @code{user} and @code{password} arguments have changed
places).
@xref{Perl DBI Class, , Perl @code{DBI} Class}.

The following changes may affect queries in old applications:

@itemize @bullet
@item
@code{HAVING} must now be specified before any @code{ORDER BY} clause.
@item
The parameters to @code{LOCATE()} have been swapped.
@item
There are some new reserved words. The most notable are @code{DATE},
@code{TIME}, and @code{TIMESTAMP}.
@end itemize


@node Upgrading-to-arch,  , Upgrading-from-3.20, Upgrade
@subsection Upgrading to Another Architecture

@cindex upgrading, different architecture

If you are using MySQL Version 3.23, you can copy the @file{.frm},
@file{.MYI}, and @file{.MYD} files between different architectures that
support the same floating-point format.  (MySQL takes care of any
byte-swapping issues.)

The MySQL @code{ISAM} data and index files (@file{.ISD} and
@file{*.ISM}, respectively) are architecture-dependent and in some cases
OS-dependent.  If you want to move your applications to another machine
that has a different architecture or OS than your current machine, you
should not try to move a database by simply copying the files to the
other machine. Use @code{mysqldump} instead.

By default, @code{mysqldump} will create a file full of SQL statements.
You can then transfer the file to the other machine and feed it as input
to the @code{mysql} client.

Try @code{mysqldump --help} to see what options are available.
If you are moving the data to a newer version of MySQL, you should use
@code{mysqldump --opt} with the newer version to get a fast, compact dump.

The easiest (although not the fastest) way to move a database between two
machines is to run the following commands on the machine on which the
database is located:

@example
shell> mysqladmin -h 'other hostname' create db_name
shell> mysqldump --opt db_name \
        | mysql -h 'other hostname' db_name
@end example

If you want to copy a database from a remote machine over a slow network,
you can use:

@example
shell> mysqladmin create db_name
shell> mysqldump -h 'other hostname' --opt --compress db_name \
        | mysql db_name
@end example

You can also store the result in a file, then transfer the file to the
target machine and load the file into the database there.  For example,
you can dump a database to a file on the source machine like this:

@example
shell> mysqldump --quick db_name | gzip > db_name.contents.gz
@end example

(The file created in this example is compressed.) Transfer the file
containing the database contents to the target machine and run these commands
there:

@example
shell> mysqladmin create db_name
shell> gunzip < db_name.contents.gz | mysql db_name
@end example

@cindex @code{mysqldump}
@cindex @code{mysqlimport}
You can also use @code{mysqldump} and @code{mysqlimport} to accomplish
the database transfer.
For big tables, this is much faster than simply using @code{mysqldump}.
In the following commands, @code{DUMPDIR} represents the full pathname
of the directory you use to store the output from @code{mysqldump}.

First, create the directory for the output files and dump the database:

@example
shell> mkdir DUMPDIR
shell> mysqldump --tab=DUMPDIR db_name
@end example

Then transfer the files in the @code{DUMPDIR} directory to some corresponding
directory on the target machine and load the files into MySQL
there:

@example
shell> mysqladmin create db_name           # create database
shell> cat DUMPDIR/*.sql | mysql db_name   # create tables in database
shell> mysqlimport db_name DUMPDIR/*.txt   # load data into tables
@end example

Also, don't forget to copy the @code{mysql} database because that's where the
grant tables (@code{user}, @code{db}, @code{host}) are stored.  You may have
to run commands as the MySQL @code{root} user on the new machine
until you have the @code{mysql} database in place.

After you import the @code{mysql} database on the new machine, execute
@code{mysqladmin flush-privileges} so that the server reloads the grant table
information.


@node Operating System Specific Notes, Perl support, Upgrade, Installing
@section Operating System Specific Notes

@menu
* Linux::                       Linux Notes (All Linux Versions)
* Windows::                     Windows Notes
* Solaris::                     Solaris Notes
* BSD Notes::                   BSD Notes
* Mac OS X::                    Mac OS X Notes
* Other Unix Notes::            Other Unix Notes
* OS/2::                        OS/2 Notes
* BeOS::                        BeOS Notes
* Novell Netware::              Novell NetWare Notes
@end menu


@node Linux, Windows, Operating System Specific Notes, Operating System Specific Notes
@subsection Linux Notes (All Linux Versions)

@menu
* Binary notes-Linux::          Linux Notes for Binary Distributions
* Linux-x86::                   Linux x86 Notes
* Linux-SPARC::                 Linux SPARC Notes
* Linux-Alpha::                 Linux Alpha Notes
* Linux-PowerPC::               Linux PowerPC Notes
* Linux-MIPS::                  Linux MIPS Notes
* Linux-IA64::                  Linux IA64 Notes
@end menu

The following notes regarding @code{glibc} apply only to the situation
when you build MySQL
yourself. If you are running Linux on an x86 machine, in most cases it is
much better for you to just use our binary. We link our binaries against
the best patched version of @code{glibc} we can come up with and with the
best compiler options, in an attempt to make it suitable for a high-load
server. So if you read the following text, and are in doubt about
what you should do, try our binary first to see if it meets your needs, and
worry about your own build only after you have discovered that our binary is
not good enough. In that case, we would appreciate a note about it, so we
can build a better binary next time. For a typical user, even for setups with
a lot of concurrent connections and/or tables exceeding 2GB limit, our
binary in most cases is the best choice.

MySQL uses LinuxThreads on Linux.  If you are using an old
Linux version that doesn't have @code{glibc2}, you must install
LinuxThreads before trying to compile MySQL.   You can get
LinuxThreads at @uref{http://www.mysql.com/Downloads/Linux/}.

@strong{Note}: we have seen some strange problems with Linux 2.2.14 and
MySQL on SMP systems. If you have a SMP system, we recommend
you upgrade to Linux 2.4 as soon as possible!  Your system will be
faster and more stable by doing this!

Note that @code{glibc} versions before and including Version 2.1.1 have
a fatal bug in @code{pthread_mutex_timedwait} handling, which is used
when you do @code{INSERT DELAYED}.  We recommend that you not use
@code{INSERT DELAYED} before upgrading glibc.

If you plan to have 1000+ concurrent connections, you will need to make
some changes to LinuxThreads, recompile it, and relink MySQL against
the new @file{libpthread.a}.  Increase @code{PTHREAD_THREADS_MAX} in
@file{sysdeps/unix/sysv/linux/bits/local_lim.h} to 4096 and decrease
@code{STACK_SIZE} in @file{linuxthreads/internals.h} to 256 KB. The paths are
relative to the root of @code{glibc} Note that MySQL will not be
stable with around 600-1000 connections if @code{STACK_SIZE} is the default
of 2 MB.

If MySQL can't open enough files,
or connections, it may be that you haven't configured Linux to handle
enough files.

In Linux 2.2 and onward, you can check the number of allocated
file handlers by doing:

@example
cat /proc/sys/fs/file-max
cat /proc/sys/fs/dquot-max
cat /proc/sys/fs/super-max
@end example

If you have more than 16 MB of memory, you should add something like the
following in your boot script (@file{/etc/rc/boot.local} on SuSE):

@example
echo 65536 > /proc/sys/fs/file-max
echo 8192 > /proc/sys/fs/dquot-max
echo 1024 > /proc/sys/fs/super-max
@end example

You can also run the preceding commands from the command-line as root, but in this case
your old limits will be used the next time your computer reboots.

You should also add /etc/my.cnf:

@example
[safe_mysqld]
open-files-limit=8192
@end example

This should allow MySQL to create up to 8192 connections + files.

The @code{STACK_SIZE} constant in LinuxThreads controls the spacing of thread
stacks in the address space.  It needs to be large enough so that there will
be plenty of room for the stack of each individual thread, but small enough
to keep the stack of some threads from running into the global @code{mysqld}
data.  Unfortunately, the Linux implementation of @code{mmap()}, as we have
experimentally discovered, will successfully unmap an already mapped region
if you ask it to map out an address already in use, zeroing out the data
on the entire page, instead of returning an error.  So, the safety of
@code{mysqld} or any other threaded application depends on the "gentleman"
behavior of the code that creates threads.  The user must take measures to
make sure the number of running threads at any time is sufficiently low for
thread stacks to stay away from the global heap.  With @code{mysqld}, you
should enforce this "gentleman" behavior by setting a reasonable value for
the @code{max_connections} variable.

If you build MySQL yourself and do not want to mess with patching
LinuxThreads, you should set @code{max_connections} to a value no higher
than 500.  It should be even less if you have a large key buffer,  large
heap tables, or some other things that make @code{mysqld} allocate a lot
of memory, or if you are running a 2.2 kernel with a 2GB patch. If you are
using our binary or RPM version 3.23.25 or later, you can safely set
@code{max_connections} at 1500, assuming no large key buffer or heap tables
with lots of data.  The more you reduce @code{STACK_SIZE} in LinuxThreads
the more threads you can safely create.  We recommend the values between
128K and 256K.

If you use a lot of concurrent connections, you may suffer from a "feature"
in the 2.2 kernel that penalises a process for forking or cloning a child
in an attempt to prevent a fork bomb attack.  This will cause MySQL
not to scale well as you increase the number of concurrent clients.  On
single-CPU systems, we have seen this manifested in a very slow thread
creation, which means it may take a long time to connect to MySQL
(as long as 1 minute), and it may take just as long to shut it down.  On
multiple-CPU systems, we have observed a gradual drop in query speed as
the number of clients increases.  In the process of trying to find a
solution, we have received a kernel patch from one of our users, who
claimed it made a lot of difference for his site.  The patch is available at
@uref{http://www.mysql.com/Downloads/Patches/linux-fork.patch}. We have
now done rather extensive testing of this patch on both development and
production systems.  It has significantly
improved @code{MySQL} performance without causing any problems and we now
recommend it to our users who are still running high-load servers on
2.2 kernels.  This issue has been fixed in the 2.4 kernel, so if you are not
satisfied with
the current performance of your system, rather than patching your 2.2 kernel,
it might be easier to just upgrade to 2.4, which will also give you a nice
SMP boost in addition to fixing this fairness bug.

We have tested MySQL on the 2.4 kernel on a 2-CPU machine and
found MySQL scales @strong{much} better@-there was virtually no slowdown
on queries throughput all the way up
to 1000 clients, and the MySQL scaling factor (computed as the ratio of
maximum throughput to the throughput with one client) was 180%.
We have observed similar results on a 4-CPU system@-virtually no
slowdown as the number of
clients was increased up to 1000, and 300% scaling factor. So for a high-load
SMP server we would definitely recommend the 2.4 kernel at this point. We
have discovered that it is essential to run @code{mysqld} process with the
highest possible priority on the 2.4 kernel to achieve maximum performance.
This can be done by adding
@code{renice -20 $$} command to @code{safe_mysqld}. In our testing on a
4-CPU machine, increasing the priority gave 60% increase in throughput with
400 clients.

We are currently also trying to collect
more info on how well @code{MySQL} performs on 2.4 kernel on 4-way and 8-way
systems. If you have access such a system and have done some benchmarks,
please send a mail to @email{docs@@mysql.com} with the results - we will
include them in the manual.

There is another issue that greatly hurts MySQL performance,
especially on SMP systems.  The implementation of mutex in
LinuxThreads in @code{glibc-2.1} is very bad for programs with many
threads that only
hold the mutex for a short time. On an SMP system, ironic as it is, if
you link MySQL against unmodified @code{LinuxThreads},
removing processors from the machine improves MySQL performance
in many cases.  We have made a patch available for @code{glibc 2.1.3}
to correct this behavior
(@uref{http://www.mysql.com/Downloads/Linux/linuxthreads-2.1-patch}).

With @code{glibc-2.2.2}
MySQL version 3.23.36 will use the adaptive mutex, which is much
better than even the patched one in @code{glibc-2.1.3}. Be warned, however,
that under some conditions, the current mutex code in @code{glibc-2.2.2}
overspins, which hurts MySQL performance. The chance of this
condition can be reduced by renicing @code{mysqld} process to the highest
priority. We have also been able to correct the overspin behavior with
a patch, available at
@uref{http://www.mysql.com/Downloads/Linux/linuxthreads-2.2.2.patch}.
It combines the correction of overspin, maximum number of
threads, and stack spacing all in one. You will need to apply it in the
@code{linuxthreads} directory with
@code{patch -p0 </tmp/linuxthreads-2.2.2.patch}.
We hope it will be included in
some form in to the future releases of @code{glibc-2.2}. In any case, if
you link against @code{glibc-2.2.2} you still need to correct
@code{STACK_SIZE} and @code{PTHREAD_THREADS_MAX}. We hope that the defaults
will be corrected to some more acceptable values for high-load
MySQL setup in the future, so that your own build can be reduced
to @code{./configure; make; make install}.

We recommend that you use the above patches to build a special static
version of @code{libpthread.a} and use it only for statically linking
against @code{MySQL}. We know that the patches are safe for @code{MySQL}
and significantly improve its performance, but we cannot say anything
about other applications. If you link other applications against the
patched version of the library, or build a patched shared version and
install it on your system, you are doing it at your own risk with regard
to other applications that depend on @code{LinuxThreads}.

If you experience any strange problems during the installation of
MySQL, or with some common utilties hanging, it is very likely that
they are either library or compiler related. If this is the case, using our
binary will resolve them.

One known problem with the binary distribution is that with older Linux
systems that use @code{libc} (like RedHat 4.x or Slackware), you will get
some non-fatal problems with hostname resolution.
@xref{Binary notes-Linux}.

When using LinuxThreads you will see a minimum of three processes
running.  These are in fact threads.  There will be one thread for the
LinuxThreads manager, one thread to handle connections, and one thread
to handle alarms and signals.

Note that the Linux kernel and the LinuxThread library can by default
only have 1024 threads.  This means that you can only have up to 1021
connections to MySQL on an unpatched system.  The page
@uref{http://www.volano.com/linuxnotes.html} contains information how to
go around this limit.

If you see a dead @code{mysqld} daemon process with @code{ps}, this usually
means that you have found a bug in MySQL or you have a corrupted
table. @xref{Crashing}.

To get a core dump on Linux if @code{mysqld} dies with a @code{SIGSEGV} signal,
you can start @code{mysqld} with the @code{--core-file} option.  Note
that you also probably need to raise the @code{core file size} by adding
@code{ulimit -c 1000000} to @code{safe_mysqld} or starting
@code{safe_mysqld} with @code{--core-file-sizes=1000000}.
@xref{safe_mysqld, , @code{safe_mysqld}}.

If you are linking your own MySQL client and get the error:

@example
ld.so.1: ./my: fatal: libmysqlclient.so.4:
open failed: No such file or directory
@end example

When executing them, the problem can be avoided by one of the following
methods:

@itemize @bullet
@item
Link the client with the following flag (instead of @code{-Lpath}):
@code{-Wl,r/path-libmysqlclient.so}.

@item
Copy @code{libmysqclient.so} to @file{/usr/lib}.

@tindex LD_RUN_PATH environment variable
@tindex environment variable, LD_RUN_PATH
@item
Add the pathname of the directory where @code{libmysqlclient.so} is located
to the @code{LD_RUN_PATH} environment variable before running your client.
@end itemize

If you are using the Fujitsu compiler @code{(fcc / FCC)} you will have
some problems compiling MySQL because the Linux header files are very
@code{gcc} oriented.

The following @code{configure} line should work with @code{fcc/FCC}:

@example
CC=fcc CFLAGS="-O -K fast -K lib -K omitfp -Kpreex -D_GNU_SOURCE \
-DCONST=const -DNO_STRTOLL_PROTO" CXX=FCC CXXFLAGS="-O -K fast -K lib \
-K omitfp -K preex --no_exceptions --no_rtti -D_GNU_SOURCE -DCONST=const \
-Dalloca=__builtin_alloca -DNO_STRTOLL_PROTO \
'-D_EXTERN_INLINE=static __inline'" ./configure --prefix=/usr/local/mysql \
--enable-assembler --with-mysqld-ldflags=-all-static --disable-shared \
--with-low-memory
@end example


@node Binary notes-Linux, Linux-x86, Linux, Linux
@subsubsection Linux Notes for Binary Distributions

@cindex binary distributions, on Linux
@cindex Linux, binary distribution

MySQL needs at least Linux Version 2.0.

@strong{Warning}:
We have reports from some MySQL users that they have got serious
stability problems with MySQL with Linux kernel 2.2.14.  If you are
using this kernel you should upgrade to 2.2.19 (or newer) or to a 2.4
kernel.  If you have a multi-cpu box, then you should seriously consider
using 2.4 as this will give you a significant speed boost.

The binary release is linked with @code{-static}, which means you do not
normally need to worry about which version of the system libraries you
have. You need not install LinuxThreads, either.  A program linked with
@code{-static} is slightly bigger than a dynamically linked program but
also slightly faster (3-5%).  One problem, however, is that you can't use
user-definable functions (UDFs) with a statically linked program.  If
you are going to write or use UDF functions (this is something only for
C or C++ programmers), you must compile MySQL yourself, using
dynamic linking.

If you are using a @code{libc}-based system (instead of a @code{glibc2}
system), you will probably get some problems with hostname resolving and
@code{getpwnam()} with the binary release. (This is because @code{glibc}
unfortunately depends on some external libraries to resolve hostnames
and @code{getpwent()}, even when compiled with @code{-static}). In this
case you probably get the following error message when you run
@code{mysql_install_db}:

@example
Sorry, the host 'xxxx' could not be looked up
@end example

or the following error when you try to run @code{mysqld} with the @code{--user}
option:

@example
getpwnam: No such file or directory
@end example

You can solve this problem in one of the following ways:

@itemize @bullet
@item
Get a MySQL source distribution (an RPM or the @code{tar.gz}
distribution) and install this instead.
@item
Execute @code{mysql_install_db --force}; this will not execute the
@code{resolveip} test in @code{mysql_install_db}.  The downside is that
you can't use host names in the grant tables; you must use IP numbers
instead (except for @code{localhost}).  If you are using an old MySQL
release that doesn't support @code{--force}, you have to remove the
@code{resolveip} test in @code{mysql_install} with an editor.
@item
Start @code{mysqld} with @code{su} instead of using @code{--user}.
@end itemize

The Linux-Intel binary and RPM releases of MySQL are configured
for the highest possible speed.  We are always trying to use the fastest
stable compiler available.

MySQL Perl support requires Version Perl 5.004_03 or newer.

On some Linux 2.2 versions, you may get the error @code{Resource
temporarily unavailable} when you do a lot of new connections to a
@code{mysqld} server over TCP/IP.

The problem is that Linux has a delay between when you close a TCP/IP
socket and until this is actually freed by the system.  As there is only
room for a finite number of TCP/IP slots, you will get the above error if
you try to do too many new TCP/IP connections during a small time, like
when you run the MySQL @file{test-connect} benchmark over
TCP/IP.

We have mailed about this problem a couple of times to different Linux
mailing lists but have never been able to resolve this properly.

The only known 'fix' to this problem is to use persistent connections in
your clients or use sockets, if you are running the database server
and clients on the same machine.  We hope that the @code{Linux 2.4}
kernel will fix this problem in the future.


@node Linux-x86, Linux-SPARC, Binary notes-Linux, Linux
@subsubsection Linux x86 Notes

MySQL requires @code{libc} Version 5.4.12 or newer. It's known to
work with @code{libc} 5.4.46.  @code{glibc} Version 2.0.6 and later should
also work. There have been some problems with the @code{glibc} RPMs from
RedHat, so if you have problems, check whether there are any updates!
The @code{glibc} 2.0.7-19 and 2.0.7-29 RPMs are known to work.

If you are using gcc 3.0 and above to compile MySQL, you must install
the @code{libstdc++v3} library before compiling MySQL; if you don't do
this you will get an error about a missing @code{__cxa_pure_virtual}
symbol during linking!

On some older Linux distributions, @code{configure} may produce an error
like this:

@example
Syntax error in sched.h. Change _P to __P in the /usr/include/sched.h file.
See the Installation chapter in the Reference Manual.
@end example

Just do what the error message says and add an extra underscore to the
@code{_P} macro that has only one underscore, then try again.

You may get some warnings when compiling; those shown here can be ignored:

@example
mysqld.cc -o objs-thread/mysqld.o
mysqld.cc: In function `void init_signals()':
mysqld.cc:315: warning: assignment of negative value `-1' to
`long unsigned int'
mysqld.cc: In function `void * signal_hand(void *)':
mysqld.cc:346: warning: assignment of negative value `-1' to
`long unsigned int'
@end example

In Debian GNU/Linux, if you want MySQL to start automatically when
the system boots, do the following:

@example
shell> cp support-files/mysql.server /etc/init.d/mysql.server
shell> /usr/sbin/update-rc.d mysql.server defaults 99
@end example

@code{mysql.server} can be found in the @file{share/mysql} directory
under the MySQL installation directory or in the
@file{support-files} directory of the MySQL source tree.

If @code{mysqld} always core dumps when it starts up, the problem may be that
you have an old @file{/lib/libc.a}.  Try renaming it, then remove
@file{sql/mysqld} and do a new @code{make install} and try again.  This
problem has been reported on some Slackware installations.

If you get the following error when linking @code{mysqld},
it means that your @file{libg++.a} is not installed correctly:

@example
/usr/lib/libc.a(putc.o): In function `_IO_putc':
putc.o(.text+0x0): multiple definition of `_IO_putc'
@end example

You can avoid using @file{libg++.a} by running @code{configure} like this:

@example
shell> CXX=gcc ./configure
@end example

@node Linux-SPARC, Linux-Alpha, Linux-x86, Linux
@subsubsection Linux SPARC Notes

In some implementations, @code{readdir_r()} is broken.  The symptom is that
@code{SHOW DATABASES} always returns an empty set.  This can be fixed by
removing @code{HAVE_READDIR_R} from @file{config.h} after configuring and
before compiling.

Some problems will require patching your Linux installation.  The patch can
be found at
@uref{http://www.mysql.com/Downloads/patches/Linux-sparc-2.0.30.diff}.
This patch is against the Linux distribution @file{sparclinux-2.0.30.tar.gz}
that is available at @code{vger.rutgers.edu} (a version of Linux that was
never merged with the official 2.0.30).  You must also install LinuxThreads
Version 0.6 or newer.


@node Linux-Alpha, Linux-PowerPC, Linux-SPARC, Linux
@subsubsection Linux Alpha Notes

MySQL Version 3.23.12 is the first MySQL version that is
tested on Linux-Alpha.  If you plan to use MySQL on Linux-Alpha,
you should ensure that you have this version or newer.

We have tested MySQL on Alpha with our benchmarks and test suite,
and it appears to work nicely.  The main thing we haven't yet had time to
test is how things works with many concurrent users.

When we compiled the standard MySQL binary we are using SuSE 6.4,
kernel 2.2.13-SMP, Compaq C compiler (V6.2-504) and Compaq C++ compiler
(V6.3-005) on a Comaq DS20 machine with an Alpha EV6 processor.

You can find the above compilers at
@uref{http://www.support.compaq.com/alpha-tools/}).  By using these compilers,
instead of gcc, we get about 9-14 % better performance with MySQL.

Note that the configure line optimised the binary for the current CPU; this
means you can only use our binary if you have an Alpha EV6 processor. We also
compile statically to avoid library problems.

@example
CC=ccc CFLAGS="-fast" CXX=cxx CXXFLAGS="-fast -noexceptions -nortti" \
./configure --prefix=/usr/local/mysql --disable-shared \
--with-extra-charsets=complex --enable-thread-safe-client \
--with-mysqld-ldflags=-non_shared --with-client-ldflags=-non_shared
@end example

If you want to use egcs the following configure line worked for us:

@example
CFLAGS="-O3 -fomit-frame-pointer" CXX=gcc \
CXXFLAGS="-O3 -fomit-frame-pointer -felide-constructors \
-fno-exceptions -fno-rtti" ./configure --prefix=/usr/local/mysql \
--disable-shared
@end example

Some known problems when running MySQL on Linux-Alpha:

@itemize @bullet
@item
Debugging threaded applications like MySQL will not work with
@code{gdb 4.18}.  You should download and use gdb 5.1 instead!

@item
If you try linking @code{mysqld} statically when using @code{gcc}, the
resulting image will core dump at start.  In other words, @strong{don't}
use @code{--with-mysqld-ldflags=-all-static} with @code{gcc}.
@end itemize


@node Linux-PowerPC, Linux-MIPS, Linux-Alpha, Linux
@subsubsection Linux PowerPC Notes

MySQL should work on MkLinux with the newest @code{glibc} package
(tested with @code{glibc} 2.0.7).


@node Linux-MIPS, Linux-IA64, Linux-PowerPC, Linux
@subsubsection Linux MIPS Notes

To get MySQL to work on Qube2, (Linux Mips), you need the
newest @code{glibc} libraries (@code{glibc-2.0.7-29C2} is known to
work).  You must also use the @code{egcs} C++ compiler
(@code{egcs-1.0.2-9}, @code{gcc 2.95.2} or newer).


@node Linux-IA64,  , Linux-MIPS, Linux
@subsubsection Linux IA64 Notes

To get MySQL to compile on Linux IA64, we use the following compile line:
Using @code{gcc-2.96}:

@example
CC=gcc CFLAGS="-O3 -fno-omit-frame-pointer" CXX=gcc \
CXXFLAGS="-O3 -fno-omit-frame-pointer -felide-constructors \
-fno-exceptions -fno-rtti" ./configure --prefix=/usr/local/mysql \
"--with-comment=Official MySQL binary" --with-extra-charsets=complex
@end example

On IA64 the MySQL client binaries are using shared libraries. This means
that if you install our binary distribution in some other place than
@file{/usr/local/mysql} you need to either modify @file{/etc/ld.so.conf}
or add the path to the directory where you have @file{libmysqlclient.so}
to the @code{LD_LIBRARY_PATH} environment variable.

@xref{Link errors}.

@node Windows, Solaris, Linux, Operating System Specific Notes
@subsection Windows Notes

This section describes using MySQL on Windows. This information
is also provided in the @file{README} file that comes with the
MySQL Windows distribution. @xref{Windows installation}.

@menu
* Win95 start::                 Starting MySQL on Windows 9x / Me
* NT start::                    Starting MySQL on Windows NT / 2000 / XP
* Windows running::             Running MySQL on Windows
* Windows and SSH::             Connecting to a remote MySQL from Windows with SSH
* Windows symbolic links::      Splitting data across different disks under Win32
* Windows client compiling::    Compiling MySQL clients on Windows.
* Windows vs Unix::             MySQL-Windows compared to Unix MySQL
@end menu


@node Win95 start, NT start, Windows, Windows
@subsubsection Starting MySQL on Windows 95, 98 or Me

MySQL uses TCP/IP to connect a client to a server. (This will
allow any machine on your network to connect to your MySQL
server.)  Because of this, you must install TCP/IP on your
machine before starting MySQL.  You can find TCP/IP on your
Windows CD-ROM.

Note that if you are using an old Windows 95 release (for example
OSR2), it's likely that you have an old Winsock package;
MySQL requires Winsock 2! You can get the newest Winsock from
@uref{http://www.microsoft.com/}.  Windows 98 has the new Winsock 2
library, so the above doesn't apply there.

To start the @code{mysqld} server, you should start an MS-DOS
window and type:

@example
C:\> C:\mysql\bin\mysqld
@end example

This will start @code{mysqld} in the background without a window.

You can kill the MySQL server by executing:

@example
C:\> C:\mysql\bin\mysqladmin -u root shutdown
@end example

This calls the MySQL administation utility as user `root', which
is the default Administrator in the MySQL grant system. Please
note that the MySQL grant system is wholly independent from any
login users under Windows.

Note that Windows 95/98/Me don't support creation of named pipes.
So on those platforms, you can only use named pipes to connect to a
remote MySQL server running on a Windows NT/2000/XP server host.
(The MySQL server must also support named pipes, of course.
For example, using @code{mysqld-opt} under NT/2000/XP will not
allow named pipe connections.  You should use either
@code{mysqld-nt} or @code{mysqld-max-nt}.)

If @code{mysqld} doesn't start, please check the
@file{\mysql\data\mysql.err} file to see if the server wrote any
message there to indicate the cause of the problem.  You can also
try to start the server with @code{mysqld --standalone}; in this
case, you may get some useful information on the screen that may
help solve the problem.

The last option is to start @code{mysqld} with
@code{--standalone --debug}.
In this case @code{mysqld} will write a log file
@file{C:\mysqld.trace} that should contain the reason why
@code{mysqld} doesn't start. @xref{Making trace files}.

Use @code{mysqld --help} to display all the options that
@code{mysqld} understands!


@node NT start, Windows running, Win95 start, Windows
@subsubsection Starting MySQL on Windows NT, 2000 or XP

To get MySQL to work with TCP/IP on Windows NT 4, you must install
service pack 3 (or newer)!

Normally you should install MySQL as a service on Windows NT/2000/XP.
In case the server was already running, first stop it using
the following command:

@example
C:\mysql\bin> mysqladmin -u root shutdown
@end example

This calls the MySQL administation utility as user `@code{root}',
which is the default @code{Administrator} in the MySQL grant system.
Please note that the MySQL grant system is wholly independent from
any login users under Windows.

Now install the server service:

@example
C:\mysql\bin> mysqld-max-nt --install
@end example

If any options are required, they must be specified as
``@code{Start parameters}'' in the Windows @code{Services}
utility before you start the MySQL service.

The @code{Services} utility
(@code{Windows Service Control Manager}) can be found in the
@code{Windows Control Panel} (under @code{Administrative Tools}
on Windows 2000). It is advisable to close the Services utility
while performing the @code{--install} or @code{--remove}
operations, this prevents some odd errors.

For information about which server binary to run, see
@ref{Windows prepare environment}.

Please note that from MySQL version 3.23.44, you have the choice
of set up the service as @code{Manual} instead (if you don't wish
the service to be started automatically during the boot process):

@example
C:\mysql\bin> mysqld-max-nt --install-manual
@end example

The service is installed with the name @code{MySQL}. Once
installed, it can be immediately started from the @code{Services}
utility, or by using the command @code{NET START MySQL}.

Once running, @code{mysqld-max-nt} can be stopped using
@code{mysqladmin}, from the Services utility or by using the
command @code{NET STOP MySQL}.

When running as a service, the operating system will automatically stop
the MySQL service on computer shutdown. In MySQL versions < 3.23.47,
Windows only waited for a few seconds for the shutdown to complete, and
killed the database server process if the time limit was exceeded
(potentially causing problems). For instance, at the next startup the
@code{InnoDB} table handler had to do crash recovery. Starting from
MySQL version 3.23.48, the Windows will wait longer for the MySQL server
shutdown to complete. If you notice this is not enough for your
intallation, it is safest to run the MySQL server not as a service, but
from the Command prompt, and shut it down with @code{mysqladmin shutdown}.

There is a problem that Windows NT (but not Windows 2000/XP) by default only
waits 20 seconds for a service to shut down, and after that kills the
service process. You can increase this default by opening the Registry
Editor @file{\winnt\system32\regedt32.exe} and editing the value of
@code{WaitToKillServiceTimeout} at
@file{HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control}
in the Registry tree. Specify the new larger value in milliseconds,
for example 120000 to have Windows NT wait upto 120 seconds.

Please note that when run as a service, @code{mysqld-max-nt}
has no access to a console and so no messages can be seen.
Errors can be checked in @file{c:\mysql\data\mysql.err}.

If you have problems installing @code{mysqld-max-nt} as a
service, try starting it with the full path:

@example
C:\> C:\mysql\bin\mysqld-max-nt --install
@end example

If this doesn't work, you can get @code{mysqld-max-nt} to
start properly by fixing the path in the registry!

If you don't want to start @code{mysqld-max-nt} as a service,
you can start it as follows:

@example
C:\> C:\mysql\bin\mysqld-max-nt --standalone
@end example

or

@example
C:\> C:\mysql\bin\mysqld --standalone --debug
@end example

The last method gives you a debug trace in
@file{C:\mysqld.trace}. @xref{Making trace files}.


@node Windows running, Windows and SSH, NT start, Windows
@subsubsection Running MySQL on Windows

@cindex TCP/IP
@cindex named pipes

MySQL supports TCP/IP on all Windows platforms and named pipes on
NT/2000/XP. The default is to use named pipes for local connections
on NT/2000/XP and TCP/IP for all other cases if the client has TCP/IP
installed. The host name specifies which protocol is used:

@multitable @columnfractions .15 .60
@item @strong{Host name} @tab @strong{Protocol}
@item NULL (none) @tab
On NT/2000/XP, try named pipes first; if that doesn't work, use TCP/IP.
On 9x/Me, TCP/IP is used.
@item . @tab Named pipes
@item localhost @tab
TCP/IP to current host
@item hostname  @tab
TCP/IP
@end multitable

You can force a MySQL client to use named pipes by specifying the
@code{--pipe} option or by specifying @code{.} as the host name.  Use the
@code{--socket} option to specify the name of the pipe.

Note that starting from 3.23.50, named pipes are only enabled if mysqld is
started with @code{--enable-named-pipe}.  This is because some users
have experienced problems shutting down the MySQL server when one uses
named pipes.

You can test whether MySQL is working by executing the
following commands:

@example
C:\> C:\mysql\bin\mysqlshow
C:\> C:\mysql\bin\mysqlshow -u root mysql
C:\> C:\mysql\bin\mysqladmin version status proc
C:\> C:\mysql\bin\mysql test
@end example

If @code{mysqld} is slow to answer to connections on Windows 9x/Me, there is
probably a problem with your DNS.  In this case, start @code{mysqld} with
@code{--skip-name-resolve} and use only @code{localhost} and IP numbers in
the MySQL grant tables.  You can also avoid DNS when connecting to a
@code{mysqld-nt} MySQL server running on NT/2000/XP by using the
@code{--pipe} argument to specify use of named pipes.  This works for most
MySQL clients.

There are two versions of the MySQL command-line tool:
@multitable @columnfractions .10 .60
@item @strong{Binary} @tab @strong{Description}
@item @code{mysql} @tab Compiled on native Windows, which offers very limited text editing capabilities.
@item @code{mysqlc} @tab Compiled with the Cygnus GNU compiler and libraries, which offers @code{readline} editing.
@end multitable

If you want to use @code{mysqlc.exe}, you must copy
@file{C:\mysql\lib\cygwinb19.dll} to your Windows system directory
(@file{\windows\system} or similar place).

The default privileges on Windows give all local users full privileges
to all databases without specifying a password.  To make MySQL
more secure, you should set a password for all users and remove the row in
the @code{mysql.user} table that has @code{Host='localhost'} and
@code{User=''}.

You should also add a password for the @code{root} user. The following
example starts by removing the anonymous user that can be used by anyone
to access the @code{test} database, then sets a @code{root} user password:

@example
C:\> C:\mysql\bin\mysql mysql
mysql> DELETE FROM user WHERE Host='localhost' AND User='';
mysql> QUIT
C:\> C:\mysql\bin\mysqladmin reload
C:\> C:\mysql\bin\mysqladmin -u root password your_password
@end example

After you've set the password, if you want to take down the @code{mysqld}
server, you can do so using this command:

@example
C:\> mysqladmin --user=root --password=your_password shutdown
@end example

If you are using the old shareware version of MySQL Version
3.21 under Windows, the above command will fail with an error:
@code{parse error near 'SET OPTION password'}.  The solution for
this is to download and upgrade to the latest MySQL version,
which is now freely available.

With the current MySQL versions you can easily add new users
and change privileges with @code{GRANT} and @code{REVOKE} commands.
@xref{GRANT}.


@node Windows and SSH, Windows symbolic links, Windows running, Windows
@subsubsection Connecting to a Remote MySQL from Windows with SSH

@c FIX this is ugly, real ugly.

@cindex SSH
@cindex connecting, remotely with SSH

Here is a note about how to connect to get a secure connection to remote
MySQL server with SSH (by David Carlson @email{dcarlson@@mplcomm.com}):

@itemize @bullet
@item
Install an SSH client on your Windows machine.  As a user, the best non-free
one I've found is from @code{SecureCRT} from @uref{http://www.vandyke.com/}.
Another option is @code{f-secure} from @uref{http://www.f-secure.com/}. You
can also find some free ones on @code{Google} at
@uref{http://directory.google.com/Top/Computers/Security/Products_and_Tools/Cryptography/SSH/Clients/Windows/}.

@item
Start your Windows SSH client.
Set @code{Host_Name = yourmysqlserver_URL_or_IP}.
Set @code{userid=your_userid} to log in to your server (probably not the same
as your MySQL login/password.

@item
Set up port forwarding. Either do a remote forward (Set @code{local_port: 3306}, @code{remote_host: yourmysqlservername_or_ip}, @code{remote_port: 3306} )
or a local forward (Set @code{port: 3306},  @code{host: localhost}, @code{remote port: 3306}).

@item
Save everything, otherwise you'll have to redo it the next time.

@item
Log in to your server with SSH session you just created.

@item
On your Windows machine, start some ODBC application (such as Access).

@item
Create a new file in Windows and link to MySQL using the ODBC
driver the same way you normally do, except type in @code{localhost}
for the MySQL host server@-not @code{yourmysqlservername}.
@end itemize

You should now have an ODBC connection to MySQL, encrypted using SSH.


@node Windows symbolic links, Windows client compiling, Windows and SSH, Windows
@subsubsection Splitting Data Across Different Disks on Windows

@cindex symbolic links
@cindex using multiple disks to start data
@cindex disks, splitting data across

Beginning with MySQL Version 3.23.16, the @code{mysqld-max}
and @code{mysql-max-nt} servers in the MySQL distribution are
compiled with the @code{-DUSE_SYMDIR} option.  This allows you to put a
database on different disk by adding a symbolic link to it
(in a manner similar to the way that symbolic links work on Unix).

On Windows, you make a symbolic link to a database by creating a file
that contains the path to the destination directory and saving this in
the @file{mysql_data} directory under the filename @file{database.sym}.
Note that the symbolic link will be used only if the directory
@file{mysql_data_dir\database} doesn't exist.

For example, if the MySQL data directory is @file{C:\mysql\data}
and you want to have database @code{foo} located at @file{D:\data\foo}, you
should create the file @file{C:\mysql\data\foo.sym} that contains the
text @code{D:\data\foo\}.  After that, all tables created in the database
@code{foo} will be created in @file{D:\data\foo}.

Note that because of the speed penalty you get when opening every table,
we have not enabled this by default even if you have compiled
MySQL with support for this. To enable symlinks you should put
in your @file{my.cnf} or @file{my.ini} file the following entry:

@example
[mysqld]
use-symbolic-links
@end example

In MySQL 4.0 we will enable symlinks by default. Then you
should instead use the @code{skip-symlink} option if you want to
disable this.


@node Windows client compiling, Windows vs Unix, Windows symbolic links, Windows
@subsubsection Compiling MySQL Clients on Windows

@cindex compiling, on Windows
@cindex Windows, compiling on

In your source files, you should include @file{windows.h} before you include
@file{mysql.h}:

@example
#if defined(_WIN32) || defined(_WIN64)
#include <windows.h>
#endif
#include <mysql.h>
@end example

You can either link your code with the dynamic @file{libmysql.lib} library,
which is just a wrapper to load in @file{libmysql.dll} on demand, or link
with the static @file{mysqlclient.lib} library.

Note that as the mysqlclient libraries are compiled as threaded libraries,
you should also compile your code to be multi-threaded!


@node Windows vs Unix,  , Windows client compiling, Windows
@subsubsection MySQL-Windows Compared to Unix MySQL

@cindex Windows, versus Unix
@cindex operating systems, Windows versus Unix

MySQL-Windows has by now proven itself to be very stable. This version
of MySQL has the same features as the corresponding Unix version
with the following exceptions:

@table @strong
@item Windows 95 and threads
Windows 95 leaks about 200 bytes of main memory for each thread creation.
Each connection in MySQL creates a new thread, so you shouldn't
run @code{mysqld} for an extended time on Windows 95 if your server handles
many connections!  Other versions of Windows don't suffer from this bug.

@item Concurrent reads
MySQL depends on the @code{pread()} and @code{pwrite()} calls to be
able to mix @code{INSERT} and @code{SELECT}.  Currently we use mutexes
to emulate @code{pread()}/@code{pwrite()}.  We will, in the long run,
replace the file level interface with a virtual interface so that we can
use the @code{readfile()}/@code{writefile()} interface on NT/2000/XP to
get more speed.
The current implementation limits the number of open files MySQL
can use to 1024, which means that you will not be able to run as many
concurrent threads on NT/2000/XP as on Unix.

@item Blocking read
MySQL uses a blocking read for each connection.
This means that:

@itemize @bullet
@item
A connection will not be disconnected automatically after 8 hours, as happens
with the Unix version of MySQL.

@item
If a connection hangs, it's impossible to break it without killing
MySQL.

@item
@code{mysqladmin kill} will not work on a sleeping connection.

@item
@code{mysqladmin shutdown} can't abort as long as there are sleeping
connections.
@end itemize

We plan to fix this problem when our Windows developers have figured out a
nice workaround.

@item @code{DROP DATABASE}
You can't drop a database that is in use by some thread.

@item Killing MySQL from the task manager
You can't kill MySQL from the task manager or with the shutdown
utility in Windows 95.  You must take it down with @code{mysqladmin shutdown}.

@item Case-insensitive names
Filenames are case-insensitive on Windows, so database and table names
are also case-insensitive in MySQL for Windows.  The only
restriction is that database and table names must be specified using the same
case throughout a given statement.  @xref{Name case sensitivity}.

@item The @samp{\} directory character
Pathname components in Windows 95 are separated by the @samp{\} character,
which is also the escape character in MySQL.  If you are using @code{LOAD
DATA INFILE} or @code{SELECT ... INTO OUTFILE}, you must double the @samp{\}
character:

@example
mysql> LOAD DATA INFILE "C:\\tmp\\skr.txt" INTO TABLE skr;
mysql> SELECT * INTO OUTFILE 'C:\\tmp\\skr.txt' FROM skr;
@end example

Alternatively, use Unix style filenames with @samp{/} characters:

@example
mysql> LOAD DATA INFILE "C:/tmp/skr.txt" INTO TABLE skr;
mysql> SELECT * INTO OUTFILE 'C:/tmp/skr.txt' FROM skr;
@end example

@item @code{Can't open named pipe} error
If you use a MySQL 3.22 version on NT with the newest mysql-clients
you will get the following error:

@example
error 2017: can't open named pipe to host: . pipe...
@end example

@tindex .my.cnf file
This is because the release version of MySQL uses named pipes on NT
by default.  You can avoid this error by using the @code{--host=localhost}
option to the new MySQL clients or create an option file
@file{C:\my.cnf} that contains the following information:

@example
[client]
host = localhost
@end example

Starting from 3.23.50, named pipes are only enabled if mysqld is started
with @code{--enable-named-pipe}.

@item @code{Access denied for user} error
If you get the error @code{Access denied for user: 'some-user@@unknown'
to database 'mysql'} when accessing a MySQL server on the same
machine, this means that MySQL can't resolve your host name
properly.

To fix this, you should create a file @file{\windows\hosts} with the
following information:

@example
127.0.0.1       localhost
@end example

@item @code{ALTER TABLE}
While you are executing an @code{ALTER TABLE} statement, the table is locked
from usage by other threads.  This has to do with the fact that on Windows,
you can't delete a file that is in use by another threads.  (In the future,
we may find some way to work around this problem.)

@item
@code{DROP TABLE} on a table that is in use by a @code{MERGE} table will
not work on Windows because @code{MERGE} handler does the table mapping
hidden from the upper layer of MySQL.  Because Windows doesn't allow you
to drop files that are open, you first must flush all @code{MERGE}
tables (with @code{FLUSH TABLES}) or drop the @code{MERGE} table before
dropping the table.  We will fix this at the same time we introduce
@code{VIEW}s.
@item
@code{DATA DIRECTORY} and @code{INDEX DIRECTORY} directives in
@code{CREATE TABLE} is ignored on Windows, because Windows doesn't support
symbolic links.
@end table

Here are some open issues for anyone who might want to help us with the Windows
release:

@cindex Windows, open issues

@itemize @bullet
@item
Make a single-user @code{MYSQL.DLL} server.  This should include everything in
a standard MySQL server, except thread creation.  This will make
MySQL much easier to use in applications that don't need a true
client/server and don't need to access the server from other hosts.

@item
Add some nice start and shutdown icons to the MySQL installation.

@item
When registering @code{mysqld} as a service with @code{--install} (on NT)
it would be nice if you could also add default options on the command-line.
For the moment, the workaround is to list the parameters in the
@file{C:\my.cnf} file instead.

@item
It would be real nice to be able to kill @code{mysqld} from the task manager.
For the moment, you must use @code{mysqladmin shutdown}.

@item
Port @code{readline} to Windows for use in the @code{mysql} command-line tool.

@item
GUI versions of the standard MySQL clients (@code{mysql},
@code{mysqlshow}, @code{mysqladmin}, and @code{mysqldump}) would be nice.

@item
It would be nice if the socket read and write functions in @file{net.c} were
interruptible. This would make it possible to kill open threads with
@code{mysqladmin kill} on Windows.

@item
@code{mysqld} always starts in the "C" locale and not in the default locale.
We would like to have @code{mysqld} use the current locale for the sort order.

@item
Add macros to use the faster thread-safe increment/decrement methods
provided by Windows.
@end itemize

Other Windows-specific issues are described in the @file{README} file that
comes with the MySQL-Windows distribution.


@node Solaris, BSD Notes, Windows, Operating System Specific Notes
@subsection Solaris Notes

@cindex Solaris installation problems
@cindex problems, installing on Solaris
@cindex tar, problems on Solaris
@cindex errors, directory checksum
@cindex checksum errors

On Solaris, you may run into trouble even before you get the MySQL
distribution unpacked!  Solaris @code{tar} can't handle long file names, so
you may see an error like this when you unpack MySQL:

@example
x mysql-3.22.12-beta/bench/Results/ATIS-mysql_odbc-NT_4.0-cmp-db2,\
informix,ms-sql,mysql,oracle,solid,sybase, 0 bytes, 0 tape blocks
tar: directory checksum error
@end example

In this case, you must use GNU @code{tar} (@code{gtar}) to unpack the
distribution.  You can find a precompiled copy for Solaris at
@uref{http://www.mysql.com/Downloads/}.

Sun native threads work only on Solaris 2.5 and higher.  For Version 2.4 and
earlier, MySQL will automatically use MIT-pthreads.
@xref{MIT-pthreads}.

If you get the following error from configure:

@example
checking for restartable system calls... configure: error can not run test
programs while cross compiling
@end example

This means that you have something wrong with your compiler installation!
In this case you should upgrade your compiler to a newer version.  You may
also be able to solve this problem by inserting the following row into the
@file{config.cache} file:

@example
ac_cv_sys_restartable_syscalls=$@{ac_cv_sys_restartable_syscalls='no'@}
@end example

If you are using Solaris on a SPARC, the recommended compiler is
@code{gcc} 2.95.2.  You can find this at @uref{http://gcc.gnu.org/}.
Note that @code{egcs} 1.1.1 and @code{gcc} 2.8.1 don't work reliably on
SPARC!

The recommended @code{configure} line when using @code{gcc} 2.95.2 is:

@example
CC=gcc CFLAGS="-O3" \
CXX=gcc CXXFLAGS="-O3 -felide-constructors -fno-exceptions -fno-rtti" \
./configure --prefix=/usr/local/mysql --with-low-memory --enable-assembler
@end example

If you have an UltraSPARC, you can get 4 % more performance by adding
"-mcpu=v8 -Wa,-xarch=v8plusa" to CFLAGS and CXXFLAGS.

If you have Sun Workshop (Fortre) 5.3 (or newer) compiler, you can
run @code{configure} like this:

@example
CC=cc CFLAGS="-Xa -fast -xO4 -native -xstrconst -mt" \
CXX=CC CXXFLAGS="-noex -xO4 -mt" \
./configure --prefix=/usr/local/mysql --enable-assembler
@end example

In the MySQL benchmarks, we got a 6 % speedup on an UltraSPARC when
using Sun Workshop 5.3 compared to using gcc with -mcpu flags.

If you get a problem with @code{fdatasync} or @code{sched_yield},
you can fix this by adding @code{LIBS=-lrt} to the configure line

The following paragraph is only relevant for older compilers than
WorkShop 5.3:

You may also have to edit the @code{configure} script to change this line:

@example
#if !defined(__STDC__) || __STDC__ != 1
@end example

to this:

@example
#if !defined(__STDC__)
@end example

If you turn on @code{__STDC__} with the @code{-Xc} option, the Sun compiler
can't compile with the Solaris @file{pthread.h} header file.  This is a Sun
bug (broken compiler or broken include file).

If @code{mysqld} issues the error message shown here when you run it, you have
tried to compile MySQL with the Sun compiler without enabling the
multi-thread option (@code{-mt}):

@example
libc internal error: _rmutex_unlock: rmutex not held
@end example

Add @code{-mt} to @code{CFLAGS} and @code{CXXFLAGS} and try again.

If you are using the SFW version of gcc (which comes with Solaris 8),
you must add @file{/opt/sfw/lib} to the environment variable
@code{LD_LIBRARY_PATH} before running configure.

If you are using the gcc available from @code{sunfreeware.com}, you may
have many problems.  You should recompile gcc and GNU binutils on the
machine you will be running them from to avoid any problems.

If you get the following error when compiling MySQL with @code{gcc},
it means that your @code{gcc} is not configured for your version of Solaris:

@example
shell> gcc -O3 -g -O2 -DDBUG_OFF  -o thr_alarm ...
./thr_alarm.c: In function `signal_hand':
./thr_alarm.c:556: too many arguments to function `sigwait'
@end example

The proper thing to do in this case is to get the newest version of
@code{gcc} and compile it with your current @code{gcc} compiler!  At
least for Solaris 2.5, almost all binary versions of @code{gcc} have
old, unusable include files that will break all programs that use
threads (and possibly other programs)!

Solaris doesn't provide static versions of all system libraries
(@code{libpthreads} and @code{libdl}), so you can't compile MySQL
with @code{--static}.  If you try to do so, you will get the error:

@example
ld: fatal: library -ldl: not found

or

undefined reference to `dlopen'

or

cannot find -lrt
@end example

If too many processes try to connect very rapidly to @code{mysqld}, you will
see this error in the MySQL log:

@example
Error in accept: Protocol error
@end example

You might try starting the server with the @code{--set-variable back_log=50}
option as a workaround for this. @xref{Command-line options}.

If you are linking your own MySQL client, you might get the
following error when you try to execute it:

@example
ld.so.1: ./my: fatal: libmysqlclient.so.#:
open failed: No such file or directory
@end example

The problem can be avoided by one of the following methods:

@itemize @bullet
@item
Link the client with the following flag (instead of @code{-Lpath}):
@code{-Wl,r/full-path-to-libmysqlclient.so}.

@item
Copy @file{libmysqclient.so} to @file{/usr/lib}.

@tindex LD_RUN_PATH environment variable
@tindex Environment variable, LD_RUN_PATH
@item
Add the pathname of the directory where @file{libmysqlclient.so} is located
to the @code{LD_RUN_PATH} environment variable before running your client.
@end itemize

If you have problems with configure trying to link with @code{-lz} and
you don't have @code{zlib} installed, you have two options:

@itemize @bullet
@item
If you want to be able to use the compressed communication protocol,
you need to get and install zlib from ftp.gnu.org.

@item
Configure with @code{--with-named-z-libs=no}.
@end itemize

If you are using gcc and have problems with loading @code{UDF} functions
into MySQL, try adding @code{-lgcc} to the link line for the
@code{UDF} function.

If you would like MySQL to start automatically, you can copy
@file{support-files/mysql.server} to @file{/etc/init.d} and create a
symbolic link to it named @file{/etc/rc3.d/S99mysql.server}.

As Solaris doesn't support core files for @code{setuid()} applications,
you can't get a core file from @code{mysqld} if you are using the
@code{--user} option.

@menu
* Solaris 2.7::                 Solaris 2.7/2.8 Notes
* Solaris x86::                 Solaris x86 Notes
@end menu


@node Solaris 2.7, Solaris x86, Solaris, Solaris
@subsubsection Solaris 2.7/2.8 Notes

You can normally use a Solaris 2.6 binary on Solaris 2.7 and 2.8.  Most
of the Solaris 2.6 issues also apply for Solaris 2.7 and 2.8.

Note that MySQL Version 3.23.4 and above should be able to autodetect
new versions of Solaris and enable workarounds for the following problems!

Solaris 2.7 / 2.8 has some bugs in the include files.  You may see the
following error when you use @code{gcc}:

@example
/usr/include/widec.h:42: warning: `getwc' redefined
/usr/include/wchar.h:326: warning: this is the location of the previous
definition
@end example

If this occurs, you can do the following to fix the problem:

Copy @code{/usr/include/widec.h} to
@code{.../lib/gcc-lib/os/gcc-version/include} and change line 41 from:

@example
#if     !defined(lint) && !defined(__lint)

to

#if     !defined(lint) && !defined(__lint) && !defined(getwc)
@end example

Alternatively, you can edit @file{/usr/include/widec.h} directly.  Either
way, after you make the fix, you should remove @file{config.cache} and run
@code{configure} again!

If you get errors like this when you run @code{make}, it's because
@code{configure} didn't detect the @file{curses.h} file (probably
because of the error in @file{/usr/include/widec.h}):

@example
In file included from mysql.cc:50:
/usr/include/term.h:1060: syntax error before `,'
/usr/include/term.h:1081: syntax error before `;'
@end example

The solution to this is to do one of the following:

@itemize @bullet
@item
Configure with @code{CFLAGS=-DHAVE_CURSES_H CXXFLAGS=-DHAVE_CURSES_H ./configure}.

@item
Edit @file{/usr/include/widec.h} as indicted above and rerun configure.

@item
Remove the @code{#define HAVE_TERM} line from @file{config.h} file and
run @code{make} again.
@end itemize

If you get a problem that your linker can't find @code{-lz} when linking
your client program, the problem is probably that your @file{libz.so} file is
installed in @file{/usr/local/lib}.  You can fix this by one of the
following methods:

@itemize @bullet
@item
Add @file{/usr/local/lib} to @code{LD_LIBRARY_PATH}.

@item
Add a link to @file{libz.so} from @file{/lib}.

@item
If you are using Solaris 8, you can install the optional zlib from your
Solaris 8 CD distribution.

@item
Configure MySQL with the @code{--with-named-z-libs=no} option.
@end itemize


@node Solaris x86,  , Solaris 2.7, Solaris
@subsubsection Solaris x86 Notes

On Solaris 2.8 on x86, @code{mysqld} will core dump if you run
'strip' in.

If you are using @code{gcc} or @code{egcs} on Solaris x86 and you
experience problems with core dumps under load, you should use the
following @code{configure} command:

@example
CC=gcc CFLAGS="-O3 -fomit-frame-pointer -DHAVE_CURSES_H" \
CXX=gcc \
CXXFLAGS="-O3 -fomit-frame-pointer -felide-constructors -fno-exceptions \
-fno-rtti -DHAVE_CURSES_H" \
./configure --prefix=/usr/local/mysql
@end example

This will avoid problems with the @code{libstdc++} library and with C++
exceptions.

If this doesn't help, you should compile a debug version and run
it with a trace file or under @code{gdb}.  @xref{Using gdb on mysqld}.


@node BSD Notes, Mac OS X, Solaris, Operating System Specific Notes
@subsection BSD Notes

This section provides information for the various BSD flavours,
as well as specific versions within those.

@menu
* FreeBSD::                     FreeBSD Notes
* NetBSD::                      NetBSD notes
* OpenBSD::                     OpenBSD 2.5 Notes
* OpenBSD 2.8::                 OpenBSD 2.8 Notes
* BSDI::                        BSD/OS Version 2.x Notes
* BSDI3::                       BSD/OS Version 3.x Notes
* BSDI4::                       BSD/OS Version 4.x Notes
@end menu


@node FreeBSD, NetBSD, BSD Notes, BSD Notes
@subsubsection FreeBSD Notes

FreeBSD 3.x is recommended for running MySQL since the thread package
is much more integrated.

The easiest and therefor the preferred way to install is to use the
mysql-server and mysql-client ports available on
@uref{http://www.freebsd.org/}.

Using these gives you:
@itemize @bullet
@item
A working MySQL with all optimisations known to work on your version
of FreeBSD enabled.

@item
Automatic configuration and build.

@item
Startup scripts installed in /usr/local/etc/rc.d.

@item
Ability to see which files that are installed with pkg_info -L. And to
remove them all with pkg_delete if you no longer want MySQL on that
machine.
@end itemize

It is recommended you use MIT-pthreads on FreeBSD 2.x and native threads on
Versions 3 and up. It is possible to run with  native threads on some late
2.2.x versions but you may encounter problems shutting down @code{mysqld}.

The MySQL @file{Makefile}s require GNU make (@code{gmake}) to work.  If
you want to compile MySQL you need to install GNU @code{make} first.

Be sure to have your name resolver setup correct.  Otherwise, you may
experience resolver delays or failures when connecting to @code{mysqld}.

Make sure that the @code{localhost} entry in the @file{/etc/hosts} file is
correct (otherwise, you will have problems connecting to the database).  The
@file{/etc/hosts} file should start with a line:

@example
127.0.0.1       localhost localhost.your.domain
@end example

The recommended way to compile and install MySQL on FreeBSD with
gcc (2.95.2 and up) is:

@example
CC=gcc CFLAGS="-O2 -fno-strength-reduce" \
CXX=gcc CXXFLAGS="-O2 -fno-rtti -fno-exceptions -felide-constructors \
-fno-strength-reduce" \
./configure --prefix=/usr/local/mysql --enable-assembler
gmake
gmake install
./scripts/mysql_install_db
cd /usr/local/mysql
./bin/mysqld_safe &
@end example

If you notice that @code{configure} will use MIT-pthreads, you should read
the MIT-pthreads notes.  @xref{MIT-pthreads}.

If you get an error from @code{make install} that it can't find
@file{/usr/include/pthreads}, @code{configure} didn't detect that you need
MIT-pthreads.  This is fixed by executing these commands:

@example
shell> rm config.cache
shell> ./configure --with-mit-threads
@end example

FreeBSD is also known to have a very low default file handle limit.
@xref{Not enough file handles}.  Uncomment the ulimit -n section in
safe_mysqld or raise the limits for the @code{mysqld} user in /etc/login.conf
(and rebuild it with cap_mkdb /etc/login.conf).  Also be sure you set the
appropriate class for this user in the password file if you are not
using the default (use: chpass mysqld-user-name).  @xref{safe_mysqld, ,
@code{safe_mysqld}}.

If you have a lot of memory you should consider rebuilding
the kernel to allow MySQL to take more than 512M of RAM.
Take a look at @code{option MAXDSIZ} in the LINT config
file for more info.

If you get problems with the current date in MySQL, setting the
@code{TZ} variable will probably help.  @xref{Environment variables}.

To get a secure and stable system you should only use FreeBSD kernels
that are marked @code{-RELEASE}.


@node NetBSD, OpenBSD, FreeBSD, BSD Notes
@subsubsection NetBSD notes

To compile on NetBSD you need GNU @code{make}. Otherwise, the compile will
crash when @code{make} tries to run @code{lint} on C++ files.


@node OpenBSD, OpenBSD 2.8, NetBSD, BSD Notes
@subsubsection OpenBSD 2.5 Notes

On OpenBSD Version 2.5, you can compile MySQL with native threads
with the following options:

@example
CFLAGS=-pthread CXXFLAGS=-pthread ./configure --with-mit-threads=no
@end example


@node OpenBSD 2.8, BSDI, OpenBSD, BSD Notes
@subsubsection OpenBSD 2.8 Notes

Our users have reported that OpenBSD 2.8 has a threading bug which causes
problems with MySQL.  The OpenBSD Developers have fixed the problem,
but as of January 25th, 2001, it's only available in the ``-current'' branch.
The symptoms of this threading bug are: slow response, high load, high CPU
usage, and crashes.

If you get an error like @code{Error in accept:: Bad file descriptor} or
error 9 when trying to open tables or directories, the problem is probably
that you haven't allocated enough file descriptors for MySQL.

In this case try starting @code{safe_mysqld} as root with the following
options:

@code{--user=mysql --open-files-limit=2048}

@node BSDI, BSDI3, OpenBSD 2.8, BSD Notes
@subsubsection BSD/OS Version 2.x Notes

If you get the following error when compiling MySQL, your
@code{ulimit} value for virtual memory is too low:

@example
item_func.h: In method `Item_func_ge::Item_func_ge(const Item_func_ge &)':
item_func.h:28: virtual memory exhausted
make[2]: *** [item_func.o] Error 1
@end example

Try using @code{ulimit -v 80000} and run @code{make} again.  If this
doesn't work and you are using @code{bash}, try switching to @code{csh}
or @code{sh}; some BSDI users have reported problems with @code{bash}
and @code{ulimit}.

If you are using @code{gcc}, you may also use have to use the
@code{--with-low-memory} flag for @code{configure} to be able to compile
@file{sql_yacc.cc}.

If you get problems with the current date in MySQL, setting the
@code{TZ} variable will probably help.  @xref{Environment variables}.


@node BSDI3, BSDI4, BSDI, BSD Notes
@subsubsection BSD/OS Version 3.x Notes

Upgrade to BSD/OS Version 3.1.  If that is not possible, install
BSDIpatch M300-038.

Use the following command when configuring MySQL:

@example
shell> env CXX=shlicc++ CC=shlicc2 \
       ./configure \
           --prefix=/usr/local/mysql \
           --localstatedir=/var/mysql \
           --without-perl \
           --with-unix-socket-path=/var/mysql/mysql.sock
@end example

The following is also known to work:

@example
shell> env CC=gcc CXX=gcc CXXFLAGS=-O3 \
       ./configure \
           --prefix=/usr/local/mysql \
           --with-unix-socket-path=/var/mysql/mysql.sock
@end example

You can change the directory locations if you wish, or just use the
defaults by not specifying any locations.

If you have problems with performance under heavy load, try using the
@code{--skip-thread-priority} option to @code{mysqld}!  This will run
all threads with the same priority; on BSDI Version 3.1, this gives better
performance (at least until BSDI fixes their thread scheduler).

If you get the error @code{virtual memory exhausted} while compiling,
you should try using @code{ulimit -v 80000} and run @code{make} again.
If this doesn't work and you are using @code{bash}, try switching to
@code{csh} or @code{sh}; some BSDI users have reported problems with
@code{bash} and @code{ulimit}.


@node BSDI4,  , BSDI3, BSD Notes
@subsubsection BSD/OS Version 4.x Notes

BSDI Version 4.x has some thread-related bugs.  If you want to use
MySQL on this, you should install all thread-related patches.  At
least M400-023 should be installed.

On some BSDI Version 4.x systems, you may get problems with shared libraries.
The symptom is that you can't execute any client programs, for example,
@code{mysqladmin}.  In this case you need to reconfigure not to use
shared libraries with the @code{--disable-shared} option to configure.

Some customers have had problems on BSDI 4.0.1 that the @code{mysqld}
binary after a while can't open tables.  This is because some
library/system related bug causes @code{mysqld} to change current
directory without asking for this!

The fix is to either upgrade to 3.23.34 or after running @code{configure}
remove the line @code{#define HAVE_REALPATH} from @code{config.h}
before running make.

Note that the above means that you can't symbolic link a database directories
to another database directory or symbolic link a table to another database
on BSDI!  (Making a symbolic link to another disk is okay).


@node Mac OS X, Other Unix Notes, BSD Notes, Operating System Specific Notes
@subsection Mac OS X Notes

@menu
* Mac OS X Public Beta::        Mac OS X Public Beta
* Mac OS X Server::             Mac OS X Server
@end menu


@node Mac OS X Public Beta, Mac OS X Server, Mac OS X, Mac OS X
@subsubsection Mac OS X Public Beta

MySQL should work without any problems on Mac OS X Public Beta
(Darwin).  You don't need the pthread patches for this OS!


@node Mac OS X Server,  , Mac OS X Public Beta, Mac OS X
@subsubsection Mac OS X Server

Before trying to configure MySQL on Mac OS X server you must
first install the pthread package from
@uref{http://www.prnet.de/RegEx/mysql.html}.

Our binary for Mac OS X is compiled on Rhapsody 5.5 with the following
configure line:

@example
CC=gcc CFLAGS="-O2 -fomit-frame-pointer" CXX=gcc CXXFLAGS="-O2 \
-fomit-frame-pointer" ./configure --prefix=/usr/local/mysql \
"--with-comment=Official MySQL binary" --with-extra-charsets=complex \
--disable-shared
@end example

You might want to also add aliases to your shell's resource file to
access @code{mysql} and @code{mysqladmin} from the command-line:

@example
alias mysql '/usr/local/mysql/bin/mysql'
alias mysqladmin '/usr/local/mysql/bin/mysqladmin'
@end example


@node Other Unix Notes, OS/2, Mac OS X, Operating System Specific Notes
@subsection Other Unix Notes

@menu
* Binary notes-HP-UX::          HP-UX Notes for Binary Distributions
* HP-UX 10.20::                 HP-UX Version 10.20 Notes
* HP-UX 11.x::                  HP-UX Version 11.x Notes
* IBM-AIX::                     IBM-AIX notes
* SunOS::                       SunOS 4 Notes
* Alpha-DEC-UNIX::              Alpha-DEC-UNIX Notes (Tru64)
* Alpha-DEC-OSF1::              Alpha-DEC-OSF/1 Notes
* SGI-Irix::                    SGI Irix Notes
* Caldera::                     Caldera Notes
* Caldera Unixware::            Caldera Unixware Version 7.0 Notes
@end menu


@node Binary notes-HP-UX, HP-UX 10.20, Other Unix Notes, Other Unix Notes
@subsubsection HP-UX Notes for Binary Distributions

@cindex HP-UX, binary distribution
@cindex binary distributions, on HP-UX

Some of the binary distributions of MySQL for HP-UX is
distributed as an HP depot file and as a tar file.  To use the depot
file you must be running at least HP-UX 10.x to have access to HP's
software depot tools.

The HP version of MySQL was compiled on an HP 9000/8xx server
under HP-UX 10.20, and uses MIT-pthreads.  It is known to work well under
this configuration.  MySQL Version 3.22.26 and newer can also be
built with HP's native thread package.

Other configurations that may work:

@itemize @bullet
@item
HP 9000/7xx running HP-UX 10.20+
@item
HP 9000/8xx running HP-UX 10.30
@end itemize

The following configurations almost definitely won't work:

@itemize @bullet
@item
HP 9000/7xx or 8xx running HP-UX 10.x where x < 2
@item
HP 9000/7xx or 8xx running HP-UX 9.x
@end itemize

To install the distribution, use one of the commands here, where
@code{/path/to/depot} is the full pathname of the depot file:

@itemize @bullet
@item
To install everything, including the server, client and development tools:

@example
shell> /usr/sbin/swinstall -s /path/to/depot mysql.full
@end example

@item
To install only the server:

@example
shell> /usr/sbin/swinstall -s /path/to/depot mysql.server
@end example

@item
To install only the client package:

@example
shell> /usr/sbin/swinstall -s /path/to/depot mysql.client
@end example

@item
To install only the development tools:

@example
shell> /usr/sbin/swinstall -s /path/to/depot mysql.developer
@end example
@end itemize

The depot places binaries and libraries in @file{/opt/mysql} and data in
@file{/var/opt/mysql}. The depot also creates the appropriate entries in
@file{/etc/init.d} and @file{/etc/rc2.d} to start the server automatically
at boot time.  Obviously, this entails being @code{root} to install.

To install the HP-UX tar.gz distribution, you must have a copy of GNU
@code{tar}.


@node HP-UX 10.20, HP-UX 11.x, Binary notes-HP-UX, Other Unix Notes
@subsubsection HP-UX Version 10.20 Notes

There are a couple of small problems when compiling MySQL on
HP-UX.  We recommend that you use @code{gcc} instead of the HP-UX native
compiler, because @code{gcc} produces better code!

We recommend using gcc 2.95 on HP-UX.  Don't use high optimisation
flags (like -O6) as this may not be safe on HP-UX.

Note that MIT-pthreads can't be compiled with the HP-UX compiler
because it can't compile @code{.S} (assembler) files.

The following configure line should work:

@example
CFLAGS="-DHPUX -I/opt/dce/include -fpic" \
CXXFLAGS="-DHPUX -I/opt/dce/include -felide-constructors -fno-exceptions \
-fno-rtti" CXX=gcc ./configure --with-pthread \
--with-named-thread-libs='-ldce' --prefix=/usr/local/mysql --disable-shared
@end example

If you are compiling @code{gcc} 2.95 yourself, you should NOT link it with
the DCE libraries (@code{libdce.a} or @code{libcma.a}) if you want to compile
MySQL with MIT-pthreads.  If you mix the DCE and MIT-pthreads
packages you will get a @code{mysqld} to which you cannot connect.  Remove
the DCE libraries while you compile @code{gcc} 2.95!


@node HP-UX 11.x, IBM-AIX, HP-UX 10.20, Other Unix Notes
@subsubsection HP-UX Version 11.x Notes

For HP-UX Version 11.x we recommend MySQL Version 3.23.15 or later.

Because of some critical bugs in the standard HP-UX libraries, you should
install the following patches before trying to run MySQL on HP-UX 11.0:

@example
PHKL_22840 Streams cumulative
PHNE_22397 ARPA cumulative
@end example

This will solve the problem of getting @code{EWOULDBLOCK} from @code{recv()}
and @code{EBADF} from @code{accept()} in threaded applications.

If you are using @code{gcc} 2.95.1 on an unpatched HP-UX 11.x system,
you will get the error:

@example
In file included from /usr/include/unistd.h:11,
                 from ../include/global.h:125,
                 from mysql_priv.h:15,
                 from item.cc:19:
/usr/include/sys/unistd.h:184: declaration of C function ...
/usr/include/sys/pthread.h:440: previous declaration ...
In file included from item.h:306,
                 from mysql_priv.h:158,
                 from item.cc:19:
@end example

The problem is that HP-UX doesn't define @code{pthreads_atfork()} consistently.
It has conflicting prototypes in
@file{/usr/include/sys/unistd.h}:184 and
@file{/usr/include/sys/pthread.h}:440 (details below).

One solution is to copy @file{/usr/include/sys/unistd.h} into
@file{mysql/include} and edit @file{unistd.h} and change it to match
the definition in @file{pthread.h}.  Here's the diff:

@example
183,184c183,184
<      extern int pthread_atfork(void (*prepare)(), void (*parent)(),
<                                                void (*child)());
---
>      extern int pthread_atfork(void (*prepare)(void), void (*parent)(void),
>                                                void (*child)(void));
@end example

After this, the following configure line should work:

@example
CFLAGS="-fomit-frame-pointer -O3 -fpic" CXX=gcc \
CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti -O3" \
./configure --prefix=/usr/local/mysql --disable-shared
@end example

Here is some information that a HP-UX Version 11.x user sent us about compiling
MySQL with HP-UX:x compiler:

@example
 Environment:
      proper compilers.
         setenv CC cc
         setenv CXX aCC
      flags
         setenv CFLAGS -D_REENTRANT
         setenv CXXFLAGS -D_REENTRANT
         setenv CPPFLAGS -D_REENTRANT
     % aCC -V
     aCC: HP ANSI C++ B3910B X.03.14.06
     % cc -V /tmp/empty.c
     cpp.ansi: HP92453-01 A.11.02.00 HP C Preprocessor (ANSI)
     ccom: HP92453-01 A.11.01.00 HP C Compiler
     cc: "/tmp/empty.c", line 1: warning 501: Empty source file.

  configuration:
     ./configure  --with-pthread        \
     --prefix=/source-control/mysql     \
     --with-named-thread-libs=-lpthread \
     --with-low-memory

    added '#define _CTYPE_INCLUDED' to include/m_ctype.h. This
    symbol is the one defined in HP's /usr/include/ctype.h:

     /* Don't include std ctype.h when this is included */
     #define _CTYPE_H
     #define __CTYPE_INCLUDED
     #define _CTYPE_INCLUDED
     #define _CTYPE_USING   /* Don't put names in global namespace. */
@end example

@itemize @bullet
@item
I had to use the compile-time flag @code{-D_REENTRANT} to get the compiler
to recognise the prototype for @code{localtime_r}. Alternatively I could have
supplied the prototype for @code{localtime_r}. But I wanted to catch other
bugs without needing to run into them.  I wasn't sure where I needed it, so I
added it to all flags.
@item
The optimisation flags used by MySQL (-O3) are not recognised by HP's
compilers.  I did not change the flags.
@end itemize

If you get the following error from @code{configure}

@example
checking for cc option to accept ANSI C... no
configure: error: MySQL requires a ANSI C compiler (and a C++ compiler).
Try gcc. See the Installation chapter in the Reference Manual.
@end example

Check that you don't have the path to the K&R compiler before the path
to the HP-UX C and C++ compiler.


@node IBM-AIX, SunOS, HP-UX 11.x, Other Unix Notes
@subsubsection IBM-AIX notes

@cindex problems, installing on IBM-AIX

Automatic detection of @code{xlC} is missing from Autoconf, so a
@code{configure} command something like this is needed when compiling
MySQL (This example uses the IBM compiler):

@example
export CC="xlc_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192 "
export CXX="xlC_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192"
export CFLAGS="-I /usr/local/include"
export LDLFAGS="-L /usr/local/lib"
export CPPFLAGS=$CFLAGS
export CXXFLAGS=$CFLAGS

./configure --prefix=/usr/local \
		--localstatedir=/var/mysql \
		--sysconfdir=/etc/mysql \
		--sbindir='/usr/local/bin' \
		--libexecdir='/usr/local/bin' \
		--enable-thread-safe-client \
		--enable-large-files
@end example

Above are the options used to compile the MySQL distribution that
can be found at @uref{http://www-frec.bull.com/}.

If you change the @code{-O3} to @code{-O2} in the above configure line,
you must also remove the @code{-qstrict} option (this is a limitation in
the IBM C compiler).

If you are using @code{gcc} or @code{egcs} to compile MySQL, you
@strong{must} use the @code{-fno-exceptions} flag, as the exception
handling in @code{gcc}/@code{egcs} is not thread-safe!  (This is tested with
@code{egcs} 1.1.)  There are also some known problems with IBM's assembler,
which may cause it to generate bad code when used with gcc.

We recommend the following @code{configure} line with @code{egcs} and
@code{gcc 2.95} on AIX:

@example
CC="gcc -pipe -mcpu=power -Wa,-many" \
CXX="gcc -pipe -mcpu=power -Wa,-many" \
CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti" \
./configure --prefix=/usr/local/mysql --with-low-memory
@end example

The @code{-Wa,-many} is necessary for the compile to be successful.  IBM is
aware of this problem but is in to hurry to fix it because of the workaround
available.  We don't know if the @code{-fno-exceptions} is required with
@code{gcc 2.95}, but as MySQL doesn't use exceptions and the above
option generates faster code, we recommend that you should always use this
option with @code{egcs / gcc}.

If you get a problem with assembler code try changing the -mcpu=xxx to
match your CPU. Typically power2, power, or powerpc may need to be used,
alternatively you might need to use 604 or 604e. I'm not positive but I
would think using "power" would likely be safe most of the time, even on
a power2 machine.

If you don't know what your CPU is then do a "uname -m", this will give
you back a string that looks like "000514676700", with a format of
xxyyyyyymmss where xx and ss are always 0's, yyyyyy is a unique system
id and mm is the id of the CPU Planar. A chart of these values can be
found at
@uref{http://publib.boulder.ibm.com/doc_link/en_US/a_doc_lib/cmds/aixcmds5/uname.htm}.
This will give you a machine type and a machine model you can use to
determine what type of CPU you have.

If you have problems with signals (MySQL dies unexpectedly
under high load) you may have found an OS bug with threads and
signals.  In this case you can tell MySQL not to use signals by
configuring with:

@example
shell> CFLAGS=-DDONT_USE_THR_ALARM CXX=gcc \
       CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti \
       -DDONT_USE_THR_ALARM" \
       ./configure --prefix=/usr/local/mysql --with-debug --with-low-memory
@end example

This doesn't affect the performance of MySQL, but has the side
effect that you can't kill clients that are ``sleeping'' on a connection with
@code{mysqladmin kill} or @code{mysqladmin shutdown}.  Instead, the client
will die when it issues its next command.

On some versions of AIX, linking with @code{libbind.a} makes
@code{getservbyname} core dump.  This is an AIX bug and should be reported
to IBM.

For AIX 4.2.1 and gcc you have to do the following changes.

After configuring, edit @file{config.h} and @file{include/my_config.h}
and change the line that says

@example
#define HAVE_SNPRINTF 1
@end example

to

@example
#undef HAVE_SNPRINTF
@end example

And finally, in @file{mysqld.cc} you need to add a prototype for initgoups.

@example
#ifdef _AIX41
extern "C" int initgroups(const char *,int);
#endif
@end example


@node SunOS, Alpha-DEC-UNIX, IBM-AIX, Other Unix Notes
@subsubsection SunOS 4 Notes

On SunOS 4, MIT-pthreads is needed to compile MySQL, which in turn
means you will need GNU @code{make}.

Some SunOS 4 systems have problems with dynamic libraries and @code{libtool}.
You can use the following @code{configure} line to avoid this problem:

@example
shell> ./configure --disable-shared --with-mysqld-ldflags=-all-static
@end example

When compiling @code{readline}, you may get warnings about duplicate defines.
These may be ignored.

When compiling @code{mysqld}, there will be some @code{implicit declaration
of function} warnings. These may be ignored.


@node Alpha-DEC-UNIX, Alpha-DEC-OSF1, SunOS, Other Unix Notes
@subsubsection Alpha-DEC-UNIX Notes (Tru64)

If you are using egcs 1.1.2 on Digital Unix, you should upgrade to gcc
2.95.2, as egcs on DEC has some serious bugs!

When compiling threaded programs under Digital Unix, the documentation
recommends using the @code{-pthread} option for @code{cc} and @code{cxx} and
the libraries @code{-lmach -lexc} (in addition to @code{-lpthread}).  You
should run @code{configure} something like this:

@example
CC="cc -pthread" CXX="cxx -pthread -O" \
./configure --with-named-thread-libs="-lpthread -lmach -lexc -lc"
@end example

When compiling @code{mysqld}, you may see a couple of warnings like this:

@example
mysqld.cc: In function void handle_connections()':
mysqld.cc:626: passing long unsigned int *' as argument 3 of
accept(int,sockadddr *, int *)'
@end example

You can safely ignore these warnings.  They occur because @code{configure}
can detect only errors, not warnings.

If you start the server directly from the command-line, you may have problems
with it dying when you log out.  (When you log out, your outstanding processes
receive a @code{SIGHUP} signal.)  If so, try starting the server like this:

@example
shell> nohup mysqld [options] &
@end example

@code{nohup} causes the command following it to ignore any @code{SIGHUP}
signal sent from the terminal.  Alternatively, start the server by running
@code{safe_mysqld}, which invokes @code{mysqld} using @code{nohup} for you.
@xref{safe_mysqld, , @code{safe_mysqld}}.

If you get a problem when compiling mysys/get_opt.c, just remove the
line #define _NO_PROTO from the start of that file!

If you are using Compac's CC compiler, the following configure line should
work:

@example
CC="cc -pthread"
CFLAGS="-O4 -ansi_alias -ansi_args -fast -inline speed all -arch host"
CXX="cxx -pthread"
CXXFLAGS="-O4 -ansi_alias -ansi_args -fast -inline speed all -arch host \
-noexceptions -nortti"
export CC CFLAGS CXX CXXFLAGS
./configure \
--prefix=/usr/local/mysql \
--with-low-memory \
--enable-large-files \
--enable-shared=yes \
--with-named-thread-libs="-lpthread -lmach -lexc -lc"
gnumake
@end example

If you get a problem with libtool, when compiling with shared libraries
as above, when linking @code{mysql}, you should be able to get around
this by issuing:

@example
cd mysql
/bin/sh ../libtool --mode=link cxx -pthread  -O3 -DDBUG_OFF \
-O4 -ansi_alias -ansi_args -fast -inline speed \
-speculate all \ -arch host  -DUNDEF_HAVE_GETHOSTBYNAME_R \
-o mysql  mysql.o readline.o sql_string.o completion_hash.o \
../readline/libreadline.a -lcurses \
../libmysql/.libs/libmysqlclient.so  -lm
cd ..
gnumake
gnumake install
scripts/mysql_install_db
@end example


@node Alpha-DEC-OSF1, SGI-Irix, Alpha-DEC-UNIX, Other Unix Notes
@subsubsection Alpha-DEC-OSF/1 Notes

If you have problems compiling and have DEC @code{CC} and @code{gcc}
installed, try running @code{configure} like this:

@example
CC=cc CFLAGS=-O CXX=gcc CXXFLAGS=-O3 \
./configure --prefix=/usr/local/mysql
@end example

If you get problems with the @file{c_asm.h} file, you can create and use
a 'dummy' @file{c_asm.h} file with:

@example
touch include/c_asm.h
CC=gcc CFLAGS=-I./include \
CXX=gcc CXXFLAGS=-O3 \
./configure --prefix=/usr/local/mysql
@end example

Note that the following problems with the @code{ld} program can be fixed
by downloading the latest DEC (Compaq) patch kit from:
@uref{http://ftp.support.compaq.com/public/unix/}.

On OSF/1 V4.0D and compiler "DEC C V5.6-071 on Digital Unix V4.0 (Rev. 878)"
the compiler had some strange behavior (undefined @code{asm} symbols).
@code{/bin/ld} also appears to be broken (problems with @code{_exit
undefined} errors occuring while linking @code{mysqld}).  On this system, we
have managed to compile MySQL with the following @code{configure}
line, after replacing @code{/bin/ld} with the version from OSF 4.0C:

@example
CC=gcc CXX=gcc CXXFLAGS=-O3 ./configure --prefix=/usr/local/mysql
@end example

With the Digital compiler "C++ V6.1-029", the following should work:

@example
CC=cc -pthread
CFLAGS=-O4 -ansi_alias -ansi_args -fast -inline speed -speculate all \
       -arch host
CXX=cxx -pthread
CXXFLAGS=-O4 -ansi_alias -ansi_args -fast -inline speed -speculate all \
          -arch host -noexceptions -nortti
export CC CFLAGS CXX CXXFLAGS
./configure --prefix=/usr/mysql/mysql --with-mysqld-ldflags=-all-static \
            --disable-shared --with-named-thread-libs="-lmach -lexc -lc"
@end example

In some versions of OSF/1, the @code{alloca()} function is broken. Fix
this by removing the line in @file{config.h} that defines @code{'HAVE_ALLOCA'}.

The @code{alloca()} function also may have an incorrect prototype in
@code{/usr/include/alloca.h}.  This warning resulting from this can be ignored.

@code{configure} will use the following thread libraries automatically:
@code{--with-named-thread-libs="-lpthread -lmach -lexc -lc"}.

When using @code{gcc}, you can also try running @code{configure} like this:

@example
shell> CFLAGS=-D_PTHREAD_USE_D4 CXX=gcc CXXFLAGS=-O3 ./configure ...
@end example

If you have problems with signals (MySQL dies unexpectedly
under high load), you may have found an OS bug with threads and
signals. In this case you can tell MySQL not to use signals by
configuring with:

@example
shell> CFLAGS=-DDONT_USE_THR_ALARM \
       CXXFLAGS=-DDONT_USE_THR_ALARM \
       ./configure ...
@end example

This doesn't affect the performance of MySQL, but has the side
effect that you can't kill clients that are ``sleeping'' on a connection with
@code{mysqladmin kill} or @code{mysqladmin shutdown}.  Instead, the client
will die when it issues its next command.

With @code{gcc} 2.95.2, you will probably run into the following compile error:

@example
sql_acl.cc:1456: Internal compiler error in `scan_region', at except.c:2566
Please submit a full bug report.
@end example

To fix this you should change to the @code{sql} directory and do a ``cut
and paste'' of the last @code{gcc} line, but change @code{-O3} to
@code{-O0} (or add @code{-O0} immediately after @code{gcc} if you don't
have any @code{-O} option on your compile line).  After this is done you
can just change back to the top-level directly and run @code{make}
again.


@node SGI-Irix, Caldera, Alpha-DEC-OSF1, Other Unix Notes
@subsubsection SGI Irix Notes

If you are using Irix Version 6.5.3 or newer @code{mysqld} will only be able to
create threads if you run it as a user with @code{CAP_SCHED_MGT}
privileges (like @code{root}) or give the @code{mysqld} server this privilege
with the following shell command:

@example
shell> chcap "CAP_SCHED_MGT+epi" /opt/mysql/libexec/mysqld
@end example

You may have to undefine some things in @file{config.h} after running
@code{configure} and before compiling.

In some Irix implementations, the @code{alloca()} function is broken.  If the
@code{mysqld} server dies on some @code{SELECT} statements, remove the lines
from @file{config.h} that define @code{HAVE_ALLOC} and @code{HAVE_ALLOCA_H}.
If @code{mysqladmin create} doesn't work, remove the line from @file{config.h}
that defines @code{HAVE_READDIR_R}.  You may have to remove the
@code{HAVE_TERM_H} line as well.

SGI recommends that you install all of the patches on this page as a set:
@uref{http://support.sgi.com/surfzone/patches/patchset/6.2_indigo.rps.html}

At the very minimum, you should install the latest kernel rollup, the
latest @code{rld} rollup, and the latest @code{libc} rollup.

You definitely need all the POSIX patches on this page, for pthreads support:

@uref{http://support.sgi.com/surfzone/patches/patchset/6.2_posix.rps.html}

If you get the something like the following error when compiling
@file{mysql.cc}:

@example
"/usr/include/curses.h", line 82: error(1084): invalid combination of type
@end example

Type the following in the top-level directory of your MySQL source
tree:

@example
shell> extra/replace bool curses_bool < /usr/include/curses.h \
> include/curses.h
shell> make
@end example

There have also been reports of scheduling problems.  If only one thread is
running, things go slow.  Avoid this by starting another client.  This may
lead to a 2-to-10-fold increase in execution speed thereafter for the other
thread.  This is a poorly understood problem with Irix threads; you may have
to improvise to find solutions until this can be fixed.

If you are compiling with @code{gcc}, you can use the following
@code{configure} command:

@example
CC=gcc CXX=gcc CXXFLAGS=-O3 \
./configure --prefix=/usr/local/mysql --enable-thread-safe-client \
--with-named-thread-libs=-lpthread
@end example

On Irix 6.5.11 with native Irix C and C++ compilers ver. 7.3.1.2, the
following is reported to work

@example
CC=cc CXX=CC CFLAGS='-O3 -n32 -TARG:platform=IP22 -I/usr/local/include \
-L/usr/local/lib' CXXFLAGS='-O3 -n32 -TARG:platform=IP22 \
-I/usr/local/include -L/usr/local/lib' ./configure \
--prefix=/usr/local/mysql --with-innodb --with-berkeley-db \
--with-libwrap=/usr/local \
--with-named-curses-libs=/usr/local/lib/libncurses.a
@end example


@node Caldera, Caldera Unixware, SGI-Irix, Other Unix Notes
@subsubsection Caldera (SCO) Notes

The current port is tested only on a ``sco3.2v5.0.4'' and
``sco3.2v5.0.5'' system.  There has also been a lot of progress on a
port to ``sco 3.2v4.2''.

For the moment the recommended compiler on OpenServer is gcc 2.95.2.  With this
you should be able to compile MySQL with just:

@example
CC=gcc CXX=gcc ./configure ... (options)
@end example

@enumerate
@item
For OpenServer 5.0.X you need to use gcc-2.95.2p1 or newer from the
Skunkware.  http://www.caldera.com/skunkware/ and choose browser
OpenServer packages or by ftp to ftp2.caldera.com in the
pub/skunkware/osr5/devtools/gcc directory.

@item
You need the port of GCC 2.5.x for this product and the Development
system.  They are required on this version of Caldera (SCO) Unix.  You cannot
just use the GCC Dev system.

@item
You should get the FSU Pthreads package and install it first.  This can be
found at
@uref{http://www.cs.wustl.edu/~schmidt/ACE_wrappers/FSU-threads.tar.gz}.
You can also get a precompiled package from
@uref{http://www.mysql.com/Downloads/SCO/FSU-threads-3.5c.tar.gz}.

@item
FSU Pthreads can be compiled with Caldera (SCO) Unix 4.2 with tcpip.  Or
OpenServer 3.0 or Open Desktop 3.0 (OS 3.0 ODT 3.0), with the Caldera (SCO)
Development System installed using a good port of GCC 2.5.x ODT or OS
3.0 you will need a good port of GCC 2.5.x There are a lot of problems
without a good port.  The port for this product requires the SCO Unix
Development system.  Without it, you are missing the libraries and the
linker that is needed.

@item
To build FSU Pthreads on your system, do the following:

@enumerate a
@item
Run @code{./configure} in the @file{threads/src} directory and select
the SCO OpenServer option. This command copies @file{Makefile.SCO5} to
@file{Makefile}.

@item
Run @code{make}.

@item
To install in the default @file{/usr/include} directory, login as root,
then @code{cd} to the @file{thread/src} directory, and run @code{make
install}.
@end enumerate

@item
Remember to use GNU @code{make} when making MySQL.

@item
If you don't start @code{safe_mysqld} as root, you probably will get only the
default 110 open files per process.  @code{mysqld} will write a note about this
in the log file.

@item
With SCO 3.2V5.0.5, you should use FSU Pthreads version 3.5c or newer.
You should also use gcc 2.95.2 or newer!

The following @code{configure} command should work:

@example
shell> ./configure --prefix=/usr/local/mysql --disable-shared
@end example

@item
With SCO 3.2V4.2, you should use FSU Pthreads version 3.5c or newer.
The following @code{configure} command should work:

@example
shell> CFLAGS="-D_XOPEN_XPG4" CXX=gcc CXXFLAGS="-D_XOPEN_XPG4" \
       ./configure \
           --prefix=/usr/local/mysql \
           --with-named-thread-libs="-lgthreads -lsocket -lgen -lgthreads" \
           --with-named-curses-libs="-lcurses"
@end example

You may get some problems with some include files. In this case, you can
find new SCO-specific include files at
@uref{http://www.mysql.com/Downloads/SCO/SCO-3.2v4.2-includes.tar.gz}.
You should unpack this file in the @file{include} directory of your
MySQL source tree.
@end enumerate

Caldera (SCO) development notes:

@itemize @bullet
@item
MySQL should automatically detect FSU Pthreads and link @code{mysqld}
with @code{-lgthreads -lsocket -lgthreads}.

@item
The Caldera (SCO) development libraries are re-entrant in FSU Pthreads.
Caldera claim sthat its libraries' functions are re-entrant, so they must
be reentrant with FSU Pthreads.  FSU Pthreads on OpenServer tries to use
the SCO scheme to make re-entrant libraries.

@item
FSU Pthreads (at least the version at @uref{http://www.mysql.com/}) comes
linked with GNU @code{malloc}.  If you encounter problems with memory usage,
make sure that @file{gmalloc.o} is included in @file{libgthreads.a} and
@file{libgthreads.so}.

@item
In FSU Pthreads, the following system calls are pthreads-aware: @code{read()},
@code{write()}, @code{getmsg()}, @code{connect()}, @code{accept()},
@code{select()}, and @code{wait()}.

@item
The CSSA-2001-SCO.35.2 (the patch is listed in custom as
erg711905-dscr_remap security patch (version 2.0.0) breaks FSU threads and
makes mysqld unstable.  You have to remove this one if you want to run
mysqld on an OpenServer 5.0.6 machine.
@end itemize

If you want to install DBI on Caldera (SCO), you have to edit the
@file{Makefile} in DBI-xxx and each subdirectory.

Note that the following assumes gcc 2.95.2 or newer:

@example
OLD:                                  NEW:
CC = cc                               CC = gcc
CCCDLFLAGS = -KPIC -W1,-Bexport       CCCDLFLAGS = -fpic
CCDLFLAGS = -wl,-Bexport              CCDLFLAGS =

LD = ld                               LD = gcc -G -fpic
LDDLFLAGS = -G -L/usr/local/lib       LDDLFLAGS = -L/usr/local/lib
LDFLAGS = -belf -L/usr/local/lib      LDFLAGS = -L/usr/local/lib

LD = ld                               LD = gcc -G -fpic
OPTIMISE = -Od                        OPTIMISE = -O1

OLD:
CCCFLAGS = -belf -dy -w0 -U M_XENIX -DPERL_SCO5 -I/usr/local/include

NEW:
CCFLAGS = -U M_XENIX -DPERL_SCO5 -I/usr/local/include
@end example

This is because the Perl dynaloader will not load the @code{DBI} modules
if they were compiled with @code{icc} or @code{cc}.

Perl works best when compiled with @code{cc}.


@node Caldera Unixware,  , Caldera, Other Unix Notes
@subsubsection Caldera (SCO) Unixware Version 7.0 Notes

You must use a version of MySQL at least as recent as Version 3.22.13
because that version fixes some portability problems under Unixware.

We have been able to compile MySQL with the following @code{configure}
command on Unixware Version 7.0.1:

@example
CC=cc CXX=CC ./configure --prefix=/usr/local/mysql
@end example

If you want to use @code{gcc}, you must use @code{gcc} 2.95.2 or newer.

Caldera provides libsocket.so.2 at
@uref{ftp://stage.caldera.com/pub/security/tools} for 
pre-OSR506 security fixes. Also, the telnetd fix at
@url{ftp://stage.caldera.com/pub/security/openserver/CSSA-2001-SCO.10/}
as both libsocket.so.2 and libresolv.so.1 with instructions for
installing on pre-OSR506 systems.

It's probably a good idea to install the above patches before trying to
compile/use MySQL.

@node OS/2, BeOS, Other Unix Notes, Operating System Specific Notes
@subsection OS/2 Notes

MySQL uses quite a few open files. Because of this, you should add
something like the following to your @file{CONFIG.SYS} file:

@example
SET EMXOPT=-c -n -h1024
@end example

If you don't do this, you will probably run into the following error:

@example
File 'xxxx' not found (Errcode: 24)
@end example

When using MySQL with OS/2 Warp 3, FixPack 29 or above is
required.  With OS/2 Warp 4, FixPack 4 or above is required. This is a
requirement of the Pthreads library.  MySQL must be installed
in a partition that supports long filenames such as HPFS, FAT32, etc.

The @file{INSTALL.CMD} script must be run from OS/2's own @file{CMD.EXE}
and may not work with replacement shells such as @file{4OS2.EXE}.

The @file{scripts/mysql-install-db} script has been renamed.  It is now called
@file{install.cmd} and is a REXX script, which will set up the default
MySQL security settings and create the WorkPlace Shell icons
for MySQL.

Dynamic module support is compiled in but not fully tested. Dynamic
modules should be compiled using the Pthreads run-time library.

@example
gcc -Zdll -Zmt -Zcrtdll=pthrdrtl -I../include -I../regex -I.. \
    -o example udf_example.cc -L../lib -lmysqlclient udf_example.def
mv example.dll example.udf
@end example

@strong{Note}: Due to limitations in OS/2, UDF module name stems must not
exceed 8 characters. Modules are stored in the @file{/mysql2/udf}
directory; the @code{safe-mysqld.cmd} script will put this directory in
the @code{BEGINLIBPATH} environment variable. When using UDF modules,
specified extensions are ignored@-it is assumed to be @file{.udf}.
For example, in Unix, the shared module might be named @file{example.so}
and you would load a function from it like this:

@example
mysql> CREATE FUNCTION metaphon RETURNS STRING SONAME "example.so";
@end example

Is OS/2, the module would be named @file{example.udf}, but you would not
specify the module extension:

@example
mysql> CREATE FUNCTION metaphon RETURNS STRING SONAME "example";
@end example


@node BeOS, Novell Netware, OS/2, Operating System Specific Notes
@subsection BeOS Notes

We are really interested in getting MySQL to work on BeOS, but
unfortunately we don't have any person who knows BeOS or has time to do
a port.

We are interested in finding someone to do a port, and we will help them
with any technical questions they may have while doing the port.

We have previously talked with some BeOS developers that have said that
MySQL is 80% ported to BeOS, but we haven't heard from them
in a while.


@node Novell Netware,  , BeOS, Operating System Specific Notes
@subsection Novell NetWare Notes

We are really interested in getting MySQL to work on NetWare, but
unfortunately we don't have any person who knows NetWare or has time to do
a port.

We are interested in finding someone to do a port, and we will help them
with any technical questions they may have while doing the port.



@node Perl support,  , Operating System Specific Notes, Installing
@section Perl Installation Comments

@cindex Perl, installing
@cindex installing, Perl

@menu
* Perl installation::           Installing Perl on Unix
* ActiveState Perl::            Installing ActiveState Perl on Windows
* Windows Perl::                Installing the MySQL Perl distribution on Windows
* Perl support problems::       Problems using the Perl @code{DBI}/@code{DBD} interface
@end menu


@node Perl installation, ActiveState Perl, Perl support, Perl support
@subsection Installing Perl on Unix

Perl support for MySQL is provided by means of the
@code{DBI}/@code{DBD} client interface.  @xref{Perl}.  The Perl
@code{DBD}/@code{DBI} client code requires Perl Version 5.004 or later.  The
interface @strong{will not work} if you have an older version of Perl.

MySQL Perl support also requires that you've installed
MySQL client programming support.  If you installed MySQL
from RPM files, client programs are in the client RPM, but client programming
support is in the developer RPM.  Make sure you've installed the latter RPM.

As of Version 3.22.8, Perl support is distributed separately from the main
MySQL distribution.  If you want to install Perl support, the files
you will need can be obtained from
@uref{http://www.mysql.com/Downloads/Contrib/}.

The Perl distributions are provided as compressed @code{tar} archives and
have names like @file{MODULE-VERSION.tar.gz}, where @code{MODULE} is the
module name and @code{VERSION} is the version number.  You should get the
@code{Data-Dumper}, @code{DBI}, and @code{Msql-Mysql-modules} distributions
and install them in that order.  The installation procedure is shown here.
The example shown is for the @code{Data-Dumper} module, but the procedure is
the same for all three distributions:

@enumerate
@item
Unpack the distribution into the current directory:
@example
shell> gunzip < Data-Dumper-VERSION.tar.gz | tar xvf -
@end example
This command creates a directory named @file{Data-Dumper-VERSION}.

@item
Change into the top-level directory of the unpacked distribution:
@example
shell> cd Data-Dumper-VERSION
@end example

@item
Build the distribution and compile everything:
@example
shell> perl Makefile.PL
shell> make
shell> make test
shell> make install
@end example
@end enumerate

The @code{make test} command is important because it verifies that the
module is working.  Note that when you run that command during the
@code{Msql-Mysql-modules} installation to exercise the interface code, the
MySQL server must be running or the test will fail.

It is a good idea to rebuild and reinstall the @code{Msql-Mysql-modules}
distribution whenever you install a new release of MySQL,
particularly if you notice symptoms such as all your @code{DBI} scripts
dumping core after you upgrade MySQL.

If you don't have the right to install Perl modules in the system directory
or if you to install local Perl modules, the following reference may help
you:

@example
@uref{http://www.iserver.com/support/contrib/perl5/modules.html}
@end example

Look under the heading
@code{Installing New Modules that Require Locally Installed Modules}.


@node ActiveState Perl, Windows Perl, Perl installation, Perl support
@subsection Installing ActiveState Perl on Windows

@cindex installing, Perl on Windows
@cindex Perl, installing on Windows
@cindex ActiveState Perl

To install the MySQL @code{DBD} module with ActiveState Perl on
Windows, you should do the following:

@itemize @bullet
@item
Get ActiveState Perl from
@uref{http://www.activestate.com/Products/ActivePerl/}
and install it.

@item
Open a DOS shell.

@item
If required, set the HTTP_proxy variable. For example, you might try:

@example
set HTTP_proxy=my.proxy.com:3128
@end example

@item
Start the PPM program:

@example
C:\> c:\perl\bin\ppm.pl
@end example

@item
If you have not already done so, install @code{DBI}:

@example
ppm> install DBI
@end example

@item
If this succeeds, run the following command:

@example
install \
ftp://ftp.de.uu.net/pub/CPAN/authors/id/JWIED/DBD-mysql-1.2212.x86.ppd
@end example
@end itemize

The above should work at least with ActiveState Perl Version 5.6.

If you can't get the above to work, you should instead install the
@code{MyODBC} driver and connect to MySQL server through
ODBC:

@example
use DBI;
$dbh= DBI->connect("DBI:ODBC:$dsn","$user","$password") ||
  die "Got error $DBI::errstr when connecting to $dsn\n";
@end example


@node Windows Perl, Perl support problems, ActiveState Perl, Perl support
@subsection Installing the MySQL Perl Distribution on Windows

The MySQL Perl distribution contains @code{DBI},
@code{DBD:MySQL} and @code{DBD:ODBC}.

@itemize @bullet
@item
Get the Perl distribution for Windows from
@uref{http://www.mysql.com/download.html}.

@item
Unzip the distribution in @code{C:} so that you get a @file{C:\PERL} directory.

@item
Add the directory @file{C:\PERL\BIN} to your path.

@item
Add the directory @file{C:\PERL\BIN\MSWIN32-x86-thread} or
@file{C:\PERL\BIN\MSWIN32-x86} to your path.

@item
Test that @code{perl} works by executing @code{perl -v} in a DOS shell.
@end itemize


@node Perl support problems,  , Windows Perl, Perl support
@subsection Problems Using the Perl @code{DBI}/@code{DBD} Interface

@cindex problems, installing Perl
@cindex Perl DBI/DBD, installation problems

If Perl reports that it can't find the @file{../mysql/mysql.so} module,
then the problem is probably that Perl can't locate the shared library
@file{libmysqlclient.so}.

You can fix this by any of the following methods:

@itemize @bullet
@item
Compile the @code{Msql-Mysql-modules} distribution with @code{perl
Makefile.PL -static -config} rather than @code{perl Makefile.PL}.

@item
Copy @code{libmysqlclient.so} to the directory where your other shared
libraries are located (probably @file{/usr/lib} or @file{/lib}).

@item
On Linux you can add the pathname of the directory where
@file{libmysqlclient.so} is located to the @file{/etc/ld.so.conf} file.

@tindex LD_RUN_PATH environment variable
@tindex Environment variable, LD_RUN_PATH
@item
Add the pathname of the directory where @file{libmysqlclient.so} is located
to the @code{LD_RUN_PATH} environment variable.
@end itemize

If you get the following errors from @code{DBD-mysql},
you are probably using @code{gcc} (or using an old binary compiled with
@code{gcc}):

@example
/usr/bin/perl: can't resolve symbol '__moddi3'
/usr/bin/perl: can't resolve symbol '__divdi3'
@end example

Add @code{-L/usr/lib/gcc-lib/... -lgcc} to the link command when the
@file{mysql.so} library gets built (check the output from @code{make} for
@file{mysql.so} when you compile the Perl client).  The @code{-L} option
should specify the pathname of the directory where @file{libgcc.a} is located
on your system.

Another cause of this problem may be that Perl and MySQL aren't both
compiled with @code{gcc}.  In this case, you can solve the mismatch by
compiling both with @code{gcc}.

If you get the following error from @code{Msql-Mysql-modules}
when you run the tests:

@example
t/00base............install_driver(mysql) failed:
Can't load '../blib/arch/auto/DBD/mysql/mysql.so' for module DBD::mysql:
../blib/arch/auto/DBD/mysql/mysql.so: undefined symbol:
uncompress at /usr/lib/perl5/5.00503/i586-linux/DynaLoader.pm line 169.
@end example

it means that you need to include the compression library, -lz, to the
link line. This can be doing the following change in the file
@file{lib/DBD/mysql/Install.pm}:

@example
$sysliblist .= " -lm";

to

$sysliblist .= " -lm -lz";
@end example

After this, you @strong{must} run 'make realclean' and then proceed with the
installation from the beginning.

If you want to use the Perl module on a system that doesn't support
dynamic linking (like Caldera/SCO) you can generate a static version of
Perl that includes @code{DBI} and @code{DBD-mysql}.  The way this works
is that you generate a version of Perl with the @code{DBI} code linked
in and install it on top of your current Perl.  Then you use that to
build a version of Perl that additionally has the @code{DBD} code linked
in, and install that.

On Caldera (SCO), you must have the following environment variables set:

@example
shell> LD_LIBRARY_PATH=/lib:/usr/lib:/usr/local/lib:/usr/progressive/lib
or
shell> LD_LIBRARY_PATH=/usr/lib:/lib:/usr/local/lib:/usr/ccs/lib:\
/usr/progressive/lib:/usr/skunk/lib
shell> LIBPATH=/usr/lib:/lib:/usr/local/lib:/usr/ccs/lib:\
/usr/progressive/lib:/usr/skunk/lib
shell> MANPATH=scohelp:/usr/man:/usr/local1/man:/usr/local/man:\
/usr/skunk/man:
@end example

First, create a Perl that includes a statically linked @code{DBI} by running
these commands in the directory where your @code{DBI} distribution is
located:

@example
shell> perl Makefile.PL -static -config
shell> make
shell> make install
shell> make perl
@end example

Then you must install the new Perl. The output of @code{make perl} will
indicate the exact @code{make} command you will need to execute to perform
the installation.  On Caldera (SCO), this is
@code{make -f Makefile.aperl inst_perl MAP_TARGET=perl}.

Next, use the just-created Perl to create another Perl that also includes a
statically-linked @code{DBD::mysql} by running these commands in the
directory where your @code{Msql-Mysql-modules} distribution is located:

@example
shell> perl Makefile.PL -static -config
shell> make
shell> make install
shell> make perl
@end example

Finally, you should install this new Perl.  Again, the output of @code{make
perl} indicates the command to use.




@node Tutorial, MySQL Database Administration, Installing, Top
@chapter Tutorial Introduction

@cindex tutorial
@cindex terminal monitor, defined
@cindex monitor, terminal
@cindex options, provided by MySQL

@menu
* Connecting-disconnecting::    Connecting to and disconnecting from the server
* Entering queries::            Entering queries
* Database use::                Creating and using a database
* Getting information::         Getting information about databases and tables
* Examples::                    Examples
* Batch mode::                  Using @code{mysql} in batch mode
* Twin::                        Queries from twin project
* Apache::                      Using MySQL with Apache
@end menu

This chapter provides a tutorial introduction to MySQL by showing
how to use the @code{mysql} client program to create and use a simple
database.  @code{mysql} (sometimes referred to as the ``terminal monitor'' or
just ``monitor'') is an interactive program that allows you to connect to a
MySQL server, run queries, and view the results.  @code{mysql} may
also be used in batch mode: you place your queries in a file beforehand, then
tell @code{mysql} to execute the contents of the file.  Both ways of using
@code{mysql} are covered here.

To see a list of options provided by @code{mysql}, invoke it with
the @code{--help} option:

@example
shell> mysql --help
@end example

This chapter assumes that @code{mysql} is installed on your machine and that
a MySQL server is available to which you can connect.  If this is
not true, contact your MySQL administrator.  (If @strong{you} are the
administrator, you will need to consult other sections of this manual.)

This chapter describes the entire process of setting up and using a
database.  If you are interested only in accessing an already-existing
database, you may want to skip over the sections that describe how to
create the database and the tables it contains.

Because this chapter is tutorial in nature, many details are necessarily left
out.  Consult the relevant sections of the manual for more
information on the topics covered here.


@node Connecting-disconnecting, Entering queries, Tutorial, Tutorial
@section Connecting to and Disconnecting from the Server

@cindex connecting, to the server
@cindex disconnecting, from the server
@cindex server, connecting
@cindex server, disconnecting

To connect to the server, you'll usually need to provide a MySQL
user name when you invoke @code{mysql} and, most likely, a password.  If the
server runs on a machine other than the one where you log in, you'll also
need to specify a hostname.  Contact your administrator to find out what
connection parameters you should use to connect (that is, what host, user name,
and password to use).  Once you know the proper parameters, you should be
able to connect like this:

@example
shell> mysql -h host -u user -p
Enter password: ********
@end example

The @code{********} represents your password; enter it when @code{mysql}
displays the @code{Enter password:} prompt.

If that works, you should see some introductory information followed by a
@code{mysql>} prompt:


@example
shell> mysql -h host -u user -p
Enter password: ********
Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is 459 to server version: 3.22.20a-log

Type 'help' for help.

mysql>
@end example

The prompt tells you that @code{mysql} is ready for you to enter commands.

Some MySQL installations allow users to connect as the anonymous
(unnamed) user to the server running on the local host.  If this is the case
on your machine, you should be able to connect to that server by invoking
@code{mysql} without any options:

@example
shell> mysql
@end example

After you have connected successfully, you can disconnect any time by typing
@code{QUIT} at the @code{mysql>} prompt:

@example
mysql> QUIT
Bye
@end example

You can also disconnect by pressing Control-D.

Most examples in the following sections assume you are connected to the
server.  They indicate this by the @code{mysql>} prompt.


@node Entering queries, Database use, Connecting-disconnecting, Tutorial
@section Entering Queries

@cindex running, queries
@cindex queries, entering
@cindex entering, queries

Make sure you are connected to the server, as discussed in the previous
section.  Doing so will not in itself select any database to work with, but
that's okay.  At this point, it's more important to find out a little about
how to issue queries than to jump right in creating tables, loading data
into them, and retrieving data from them.  This section describes the basic
principles of entering commands, using several queries you can try out to
familiarise yourself with how @code{mysql} works.

Here's a simple command that asks the server to tell you its version number
and the current date.  Type it in as shown here following the @code{mysql>}
prompt and press Enter:

@example
mysql> SELECT VERSION(), CURRENT_DATE;
+--------------+--------------+
| VERSION()    | CURRENT_DATE |
+--------------+--------------+
| 3.22.20a-log | 1999-03-19   |
+--------------+--------------+
1 row in set (0.01 sec)
mysql>
@end example

This query illustrates several things about @code{mysql}:

@itemize @bullet
@item
A command normally consists of a SQL statement followed by a semicolon.
(There are some exceptions where a semicolon is not needed.  @code{QUIT},
mentioned earlier, is one of them.  We'll get to others later.)

@item
When you issue a command, @code{mysql} sends it to the server for execution
and displays the results, then prints another @code{mysql>} to indicate
that it is ready for another command.

@item
@code{mysql} displays query output as a table (rows and columns).  The first
row contains labels for the columns.  The rows following are the query
results.  Normally, column labels are the names of the columns you fetch from
database tables.  If you're retrieving the value of an expression rather than
a table column (as in the example just shown), @code{mysql} labels the column
using the expression itself.

@item
@code{mysql} shows how many rows were returned and how long the query took
to execute, which gives you a rough idea of server performance.  These values
are imprecise because they represent wall clock time (not CPU or machine
time), and because they are affected by factors such as server load and
network latency.  (For brevity, the ``rows in set'' line is not shown in
the remaining examples in this chapter.)
@end itemize

Keywords may be entered in any lettercase.  The following queries are
equivalent:

@example
mysql> SELECT VERSION(), CURRENT_DATE;
mysql> select version(), current_date;
mysql> SeLeCt vErSiOn(), current_DATE;
@end example

Here's another query.  It demonstrates that you can use @code{mysql} as a
simple calculator:

@example
mysql> SELECT SIN(PI()/4), (4+1)*5;
+-------------+---------+
| SIN(PI()/4) | (4+1)*5 |
+-------------+---------+
|    0.707107 |      25 |
+-------------+---------+
@end example

The commands shown thus far have been relatively short, single-line
statements.  You can even enter multiple statements on a single line.
Just end each one with a semicolon:

@example
mysql> SELECT VERSION(); SELECT NOW();
+--------------+
| VERSION()    |
+--------------+
| 3.22.20a-log |
+--------------+

+---------------------+
| NOW()               |
+---------------------+
| 1999-03-19 00:15:33 |
+---------------------+
@end example

A command need not be given all on a single line, so lengthy commands that
require several lines are not a problem.  @code{mysql} determines where your
statement ends by looking for the terminating semicolon, not by looking for
the end of the input line.  (In other words, @code{mysql}
accepts free-format input:  it collects input lines but does not execute them
until it sees the semicolon.)

Here's a simple multiple-line statement:

@example
mysql> SELECT
    -> USER()
    -> ,
    -> CURRENT_DATE;
+--------------------+--------------+
| USER()             | CURRENT_DATE |
+--------------------+--------------+
| joesmith@@localhost | 1999-03-18   |
+--------------------+--------------+
@end example

In this example, notice how the prompt changes from @code{mysql>} to
@code{->} after you enter the first line of a multiple-line query.  This is
how @code{mysql} indicates that it hasn't seen a complete statement and is
waiting for the rest.  The prompt is your friend, because it provides
valuable feedback.  If you use that feedback, you will always be aware of
what @code{mysql} is waiting for.

If you decide you don't want to execute a command that you are in the
process of entering, cancel it by typing @code{\c}:

@example
mysql> SELECT
    -> USER()
    -> \c
mysql>
@end example

Here, too, notice the prompt.  It switches back to @code{mysql>} after you
type @code{\c}, providing feedback to indicate that @code{mysql} is ready
for a new command.

The following table shows each of the prompts you may see and summarises what
they mean about the state that @code{mysql} is in:

@cindex prompts, meanings
@multitable @columnfractions .10 .80
@item @strong{Prompt} @tab @strong{Meaning}
@item @code{mysql>} @tab Ready for new command.
@item @code{@ @ @ @ ->} @tab Waiting for next line of multiple-line command.
@item @code{@ @ @ @ '>} @tab Waiting for next line, collecting a string that begins
with a single quote (@samp{'}).
@item @code{@ @ @ @ ">} @tab Waiting for next line, collecting a string that begins
with a double quote (@samp{"}).
@end multitable

Multiple-line statements commonly occur by accident when you intend to
issue a command on a single line, but forget the terminating semicolon.  In
this case, @code{mysql} waits for more input:

@example
mysql> SELECT USER()
    ->
@end example

If this happens to you (you think you've entered a statement but the only
response is a @code{->} prompt), most likely @code{mysql} is waiting for the
semicolon.  If you don't notice what the prompt is telling you, you might sit
there for a while before realising what you need to do.  Enter a semicolon to
complete the statement, and @code{mysql} will execute it:

@example
mysql> SELECT USER()
    -> ;
+--------------------+
| USER()             |
+--------------------+
| joesmith@@localhost |
+--------------------+
@end example

The @code{'>} and @code{">} prompts occur during string collection.
In MySQL, you can write strings surrounded by either @samp{'}
or @samp{"} characters (for example, @code{'hello'} or @code{"goodbye"}),
and @code{mysql} lets you enter strings that span multiple lines.  When you
see a @code{'>} or @code{">} prompt, it means that you've entered a line
containing a string that begins with a @samp{'} or @samp{"} quote character,
but have not yet entered the matching quote that terminates the string.
That's fine if you really are entering a multiple-line string, but how likely
is that?  Not very.  More often, the @code{'>} and @code{">} prompts indicate
that you've inadvertantly left out a quote character.  For example:

@example
mysql> SELECT * FROM my_table WHERE name = "Smith AND age < 30;
    ">
@end example

If you enter this @code{SELECT} statement, then press Enter and wait for the
result, nothing will happen.  Instead of wondering why this
query takes so long, notice the clue provided by the @code{">} prompt.  It
tells you that @code{mysql} expects to see the rest of an unterminated
string.  (Do you see the error in the statement?  The string @code{"Smith} is
missing the second quote.)

At this point, what do you do?  The simplest thing is to cancel the command.
However, you cannot just type @code{\c} in this case, because @code{mysql}
interprets it as part of the string that it is collecting!  Instead, enter
the closing quote character (so @code{mysql} knows you've finished the
string), then type @code{\c}:

@example
mysql> SELECT * FROM my_table WHERE name = "Smith AND age < 30;
    "> "\c
mysql>
@end example

The prompt changes back to @code{mysql>}, indicating that @code{mysql}
is ready for a new command.

It's important to know what the @code{'>} and @code{">} prompts signify,
because if you mistakenly enter an unterminated string, any further lines you
type will appear to be ignored by @code{mysql}@-including a line
containing @code{QUIT}!  This can be quite confusing, especially if you
don't know that you need to supply the terminating quote before you can
cancel the current command.


@node Database use, Getting information, Entering queries, Tutorial
@section Creating and Using a Database

@cindex databases, creating
@cindex databases, using
@cindex creating, databases

@menu
* Creating database::           Creating a database
* Creating tables::             Creating a table
* Loading tables::              Loading data into a table
* Retrieving data::             Retrieving information from a table
@end menu

Now that you know how to enter commands, it's time to access a database.

Suppose you have several pets in your home (your menagerie) and you'd
like to keep track of various types of information about them.  You can do so
by creating tables to hold your data and loading them with the desired
information.  Then you can answer different sorts of questions about your
animals by retrieving data from the tables.  This section shows you how to:

@itemize @bullet
@item
Create a database
@item
Create a table
@item
Load data into the table
@item
Retrieve data from the table in various ways
@item
Use multiple tables
@end itemize

The menagerie database will be simple (deliberately), but it is not difficult
to think of real-world situations in which a similar type of database might
be used.  For example, a database like this could be used by a farmer to keep
track of livestock, or by a veterinarian to keep track of patient records.
A menagerie distribution containing some of the queries and sample data used
in the following sections can be obtained from the MySQL web site.
It's available in either compressed @code{tar} format
(@uref{http://www.mysql.com/Downloads/Contrib/Examples/menagerie.tar.gz})
or Zip format
(@uref{http://www.mysql.com/Downloads/Contrib/Examples/menagerie.zip}).

Use the @code{SHOW} statement to find out what databases currently exist
on the server:

@example
mysql> SHOW DATABASES;
+----------+
| Database |
+----------+
| mysql    |
| test     |
| tmp      |
+----------+
@end example

The list of databases is probably different on your machine, but the
@code{mysql} and @code{test} databases are likely to be among them.  The
@code{mysql} database is required because it describes user access
privileges.  The @code{test} database is often provided as a workspace for
users to try things out.

If the @code{test} database exists, try to access it:

@example
mysql> USE test
Database changed
@end example

Note that @code{USE}, like @code{QUIT}, does not require a semicolon.  (You
can terminate such statements with a semicolon if you like; it does no harm.)
The @code{USE} statement is special in another way, too:  it must be given on
a single line.

You can use the @code{test} database (if you have access to it) for the
examples that follow, but anything you create in that database can be
removed by anyone else with access to it.  For this reason, you should
probably ask your MySQL administrator for permission to use a
database of your own.  Suppose you want to call yours @code{menagerie}.  The
administrator needs to execute a command like this:

@example
mysql> GRANT ALL ON menagerie.* TO your_mysql_name;
@end example

where @code{your_mysql_name} is the MySQL user name assigned to
you.


@node Creating database, Creating tables, Database use, Database use
@subsection Creating and Selecting a Database

@cindex selecting, databases
@cindex databases, selecting

If the administrator creates your database for you when setting up your
permissions, you can begin using it.  Otherwise, you need to create it
yourself:

@example
mysql> CREATE DATABASE menagerie;
@end example

Under Unix, database names are case-sensitive (unlike SQL keywords), so you
must always refer to your database as @code{menagerie}, not as
@code{Menagerie}, @code{MENAGERIE}, or some other variant.  This is also true
for table names.  (Under Windows, this restriction does not apply, although
you must refer to databases and tables using the same lettercase throughout a
given query.)

Creating a database does not select it for use; you must do that explicitly.
To make @code{menagerie} the current database, use this command:

@example
mysql> USE menagerie
Database changed
@end example

Your database needs to be created only once, but you must select it for use
each time you begin a @code{mysql} session.  You can do this by issuing a
@code{USE} statement as shown above.  Alternatively, you can select the
database on the command-line when you invoke @code{mysql}.  Just specify its
name after any connection parameters that you might need to provide.  For
example:

@example
shell> mysql -h host -u user -p menagerie
Enter password: ********
@end example

Note that @code{menagerie} is not your password on the command just shown.
If you want to supply your password on the command-line after the @code{-p}
option, you must do so with no intervening space (for example, as
@code{-pmypassword}, not as @code{-p mypassword}).  However, putting your
password on the command-line is not recommended, because doing so exposes it
to snooping by other users logged in on your machine.


@node Creating tables, Loading tables, Creating database, Database use
@subsection Creating a Table

@cindex tables, creating
@cindex creating, tables

Creating the database is the easy part, but at this point it's empty, as
@code{SHOW TABLES} will tell you:

@example
mysql> SHOW TABLES;
Empty set (0.00 sec)
@end example

The harder part is deciding what the structure of your database should be:
what tables you will need and what columns will be in each of them.

You'll want a table that contains a record for each of your pets.  This can
be called the @code{pet} table, and it should contain, as a bare minimum,
each animal's name.  Because the name by itself is not very interesting, the
table should contain other information.  For example, if more than one person
in your family keeps pets, you might want to list each animal's owner.  You
might also want to record some basic descriptive information such as species
and sex.

How about age?  That might be of interest, but it's not a good thing to store
in a database.  Age changes as time passes, which means you'd have to update
your records often.  Instead, it's better to store a fixed value such as
date of birth.  Then, whenever you need age, you can calculate it as the
difference between the current date and the birth date.  MySQL
provides functions for doing date arithmetic, so this is not difficult.
Storing birth date rather than age has other advantages, too:

@itemize @bullet
@item
You can use the database for tasks such as generating reminders for upcoming
pet birthdays.  (If you think this type of query is somewhat silly, note that
it is the same question you might ask in the context of a business database
to identify clients to whom you'll soon need to send out birthday greetings,
for that computer-assisted personal touch.)

@item
You can calculate age in relation to dates other than the current date.  For
example, if you store death date in the database, you can easily calculate
how old a pet was when it died.
@end itemize

You can probably think of other types of information that would be useful in
the @code{pet} table, but the ones identified so far are sufficient for now:
name, owner, species, sex, birth, and death.

Use a @code{CREATE TABLE} statement to specify the layout of your table:

@example
mysql> CREATE TABLE pet (name VARCHAR(20), owner VARCHAR(20),
    -> species VARCHAR(20), sex CHAR(1), birth DATE, death DATE);
@end example

@code{VARCHAR} is a good choice for the @code{name}, @code{owner}, and
@code{species} columns because the column values will vary in length.  The
lengths of those columns need not all be the same, and need not be
@code{20}.  You can pick any length from @code{1} to @code{255}, whatever
seems most reasonable to you.  (If you make a poor choice and it turns
out later that you need a longer field, MySQL provides an
@code{ALTER TABLE} statement.)

Animal sex can be represented in a variety of ways, for example, @code{"m"}
and @code{"f"}, or perhaps @code{"male"} and @code{"female"}.  It's simplest
to use the single characters @code{"m"} and @code{"f"}.

The use of the @code{DATE} data type for the @code{birth} and @code{death}
columns is a fairly obvious choice.

Now that you have created a table, @code{SHOW TABLES} should produce some
output:

@example
mysql> SHOW TABLES;
+---------------------+
| Tables in menagerie |
+---------------------+
| pet                 |
+---------------------+
@end example

To verify that your table was created the way you expected, use
a @code{DESCRIBE} statement:

@example
mysql> DESCRIBE pet;
+---------+-------------+------+-----+---------+-------+
| Field   | Type        | Null | Key | Default | Extra |
+---------+-------------+------+-----+---------+-------+
| name    | varchar(20) | YES  |     | NULL    |       |
| owner   | varchar(20) | YES  |     | NULL    |       |
| species | varchar(20) | YES  |     | NULL    |       |
| sex     | char(1)     | YES  |     | NULL    |       |
| birth   | date        | YES  |     | NULL    |       |
| death   | date        | YES  |     | NULL    |       |
+---------+-------------+------+-----+---------+-------+
@end example

You can use @code{DESCRIBE} any time, for example, if you forget the names of
the columns in your table or what types they are.


@node Loading tables, Retrieving data, Creating tables, Database use
@subsection Loading Data into a Table

@cindex loading, tables
@cindex tables, loading data
@cindex data, loading into tables

After creating your table, you need to populate it.  The @code{LOAD DATA} and
@code{INSERT} statements are useful for this.

Suppose your pet records can be described as shown here.
(Observe that MySQL expects dates in @code{YYYY-MM-DD} format;
this may be different from what you are used to.)

@multitable @columnfractions .10 .10 .10 .05 .15 .15
@item @strong{name} @tab @strong{owner} @tab @strong{species} @tab @strong{sex} @tab @strong{birth} @tab @strong{death}
@item Fluffy @tab Harold @tab cat @tab f @tab 1993-02-04 @tab
@item Claws @tab Gwen @tab cat @tab m @tab 1994-03-17 @tab
@item Buffy @tab Harold @tab dog @tab f @tab 1989-05-13 @tab
@item Fang @tab Benny @tab dog @tab m @tab 1990-08-27 @tab
@item Bowser @tab Diane @tab dog @tab m @tab 1998-08-31 @tab 1995-07-29
@item Chirpy @tab Gwen @tab bird @tab f @tab 1998-09-11 @tab
@item Whistler @tab Gwen @tab bird @tab @tab 1997-12-09 @tab
@item Slim @tab Benny @tab snake @tab m @tab 1996-04-29 @tab
@end multitable

Because you are beginning with an empty table, an easy way to populate it is to
create a text file containing a row for each of your animals, then load the
contents of the file into the table with a single statement.

You could create a text file @file{pet.txt} containing one record per line,
with values separated by tabs, and given in the order in which the columns
were listed in the @code{CREATE TABLE} statement.  For missing values (such
as unknown sexes or death dates for animals that are still living), you can
use @code{NULL} values.  To represent these in your text file, use
@code{\N}.  For example, the record for Whistler the bird would look like
this (where the whitespace between values is a single tab character):

@multitable @columnfractions .10 .10 .10 .05 .15 .15
@item @strong{name} @tab @strong{owner} @tab @strong{species} @tab @strong{sex} @tab @strong{birth} @tab @strong{death}
@item @code{Whistler} @tab @code{Gwen} @tab @code{bird} @tab @code{\N} @tab @code{1997-12-09} @tab @code{\N}
@end multitable

To load the text file @file{pet.txt} into the @code{pet} table, use this
command:

@example
mysql> LOAD DATA LOCAL INFILE "pet.txt" INTO TABLE pet;
@end example

You can specify the column value separator and end of line marker explicitly
in the @code{LOAD DATA} statement if you wish, but the defaults are tab and
linefeed.  These are sufficient for the statement to read the file
@file{pet.txt} properly.

When you want to add new records one at a time, the @code{INSERT} statement
is useful.  In its simplest form, you supply values for each column, in the
order in which the columns were listed in the @code{CREATE TABLE} statement.
Suppose Diane gets a new hamster named Puffball.  You could add a new record
using an @code{INSERT} statement like this:

@example
mysql> INSERT INTO pet
    -> VALUES ('Puffball','Diane','hamster','f','1999-03-30',NULL);
@end example

Note that string and date values are specified as quoted strings here.  Also,
with @code{INSERT}, you can insert @code{NULL} directly to represent a
missing value.  You do not use @code{\N} like you do with @code{LOAD DATA}.

From this example, you should be able to see that there would be a lot more
typing involved to load
your records initially using several @code{INSERT} statements rather
than a single @code{LOAD DATA} statement.


@node Retrieving data,  , Loading tables, Database use
@subsection Retrieving Information from a Table

@cindex data, retrieving
@cindex tables, retrieving data
@cindex retrieving, data from tables
@cindex unloading, tables

@menu
* Selecting all::               Selecting all data
* Selecting rows::              Selecting particular rows
* Selecting columns::           Selecting particular columns
* Sorting rows::                Sorting rows
* Date calculations::           Date calculations
* Working with NULL::           Working with @code{NULL} values
* Pattern matching::            Pattern matching
* Counting rows::               Counting rows
* Multiple tables::             Using More Than one Table
@end menu

The @code{SELECT} statement is used to pull information from a table.
The general form of the statement is:

@example
SELECT what_to_select
FROM which_table
WHERE conditions_to_satisfy
@end example

@code{what_to_select} indicates what you want to see.  This can be a list of
columns, or @code{*} to indicate ``all columns.'' @code{which_table}
indicates the table from which you want to retrieve data.  The @code{WHERE}
clause is optional.  If it's present, @code{conditions_to_satisfy} specifies
conditions that rows must satisfy to qualify for retrieval.


@node Selecting all, Selecting rows, Retrieving data, Retrieving data
@subsubsection Selecting All Data

The simplest form of @code{SELECT} retrieves everything from a table:

@example
mysql> SELECT * FROM pet;
+----------+--------+---------+------+------------+------------+
| name     | owner  | species | sex  | birth      | death      |
+----------+--------+---------+------+------------+------------+
| Fluffy   | Harold | cat     | f    | 1993-02-04 | NULL       |
| Claws    | Gwen   | cat     | m    | 1994-03-17 | NULL       |
| Buffy    | Harold | dog     | f    | 1989-05-13 | NULL       |
| Fang     | Benny  | dog     | m    | 1990-08-27 | NULL       |
| Bowser   | Diane  | dog     | m    | 1998-08-31 | 1995-07-29 |
| Chirpy   | Gwen   | bird    | f    | 1998-09-11 | NULL       |
| Whistler | Gwen   | bird    | NULL | 1997-12-09 | NULL       |
| Slim     | Benny  | snake   | m    | 1996-04-29 | NULL       |
| Puffball | Diane  | hamster | f    | 1999-03-30 | NULL       |
+----------+--------+---------+------+------------+------------+
@end example

This form of @code{SELECT} is useful if you want to review your entire table,
for instance, after you've just loaded it with your initial dataset.  As it
happens, the output just shown reveals an error in your datafile:  Bowser
appears to have been born after he died!  Consulting your original pedigree
papers, you find that the correct birth year is 1989, not 1998.

There are are least a couple of ways to fix this:

@itemize @bullet
@item
Edit the file @file{pet.txt} to correct the error, then empty the table
and reload it using @code{DELETE} and @code{LOAD DATA}:

@example
mysql> SET AUTOCOMMIT=1;  # Used for quick re-create of the table
mysql> DELETE FROM pet;
mysql> LOAD DATA LOCAL INFILE "pet.txt" INTO TABLE pet;
@end example

However, if you do this, you must also re-enter the record for Puffball.

@item
Fix only the erroneous record with an @code{UPDATE} statement:

@example
mysql> UPDATE pet SET birth = "1989-08-31" WHERE name = "Bowser";
@end example
@end itemize

As shown above, it is easy to retrieve an entire table.  But typically you
don't want to do that, particularly when the table becomes large.  Instead,
you're usually more interested in answering a particular question, in which
case you specify some constraints on the information you want.  Let's look at
some selection queries in terms of questions about your pets that they
answer.


@node Selecting rows, Selecting columns, Selecting all, Retrieving data
@subsubsection Selecting Particular Rows

@cindex rows, selecting
@cindex tables, selecting rows

You can select only particular rows from your table.  For example, if you want
to verify the change that you made to Bowser's birth date, select Bowser's
record like this:

@example
mysql> SELECT * FROM pet WHERE name = "Bowser";
+--------+-------+---------+------+------------+------------+
| name   | owner | species | sex  | birth      | death      |
+--------+-------+---------+------+------------+------------+
| Bowser | Diane | dog     | m    | 1989-08-31 | 1995-07-29 |
+--------+-------+---------+------+------------+------------+
@end example

The output confirms that the year is correctly recorded now as 1989, not 1998.

String comparisons are normally case-insensitive, so you can specify the
name as @code{"bowser"}, @code{"BOWSER"}, etc.  The query result will be
the same.

You can specify conditions on any column, not just @code{name}.  For example,
if you want to know which animals were born after 1998, test the @code{birth}
column:

@example
mysql> SELECT * FROM pet WHERE birth >= "1998-1-1";
+----------+-------+---------+------+------------+-------+
| name     | owner | species | sex  | birth      | death |
+----------+-------+---------+------+------------+-------+
| Chirpy   | Gwen  | bird    | f    | 1998-09-11 | NULL  |
| Puffball | Diane | hamster | f    | 1999-03-30 | NULL  |
+----------+-------+---------+------+------------+-------+
@end example

You can combine conditions, for example, to locate female dogs:

@example
mysql> SELECT * FROM pet WHERE species = "dog" AND sex = "f";
+-------+--------+---------+------+------------+-------+
| name  | owner  | species | sex  | birth      | death |
+-------+--------+---------+------+------------+-------+
| Buffy | Harold | dog     | f    | 1989-05-13 | NULL  |
+-------+--------+---------+------+------------+-------+
@end example

The preceding query uses the @code{AND} logical operator.  There is also an
@code{OR} operator:

@example
mysql> SELECT * FROM pet WHERE species = "snake" OR species = "bird";
+----------+-------+---------+------+------------+-------+
| name     | owner | species | sex  | birth      | death |
+----------+-------+---------+------+------------+-------+
| Chirpy   | Gwen  | bird    | f    | 1998-09-11 | NULL  |
| Whistler | Gwen  | bird    | NULL | 1997-12-09 | NULL  |
| Slim     | Benny | snake   | m    | 1996-04-29 | NULL  |
+----------+-------+---------+------+------------+-------+
@end example

@code{AND} and @code{OR} may be intermixed.  If you do that, it's a good idea
to use parentheses to indicate how conditions should be grouped:

@example
mysql> SELECT * FROM pet WHERE (species = "cat" AND sex = "m")
    -> OR (species = "dog" AND sex = "f");
+-------+--------+---------+------+------------+-------+
| name  | owner  | species | sex  | birth      | death |
+-------+--------+---------+------+------------+-------+
| Claws | Gwen   | cat     | m    | 1994-03-17 | NULL  |
| Buffy | Harold | dog     | f    | 1989-05-13 | NULL  |
+-------+--------+---------+------+------------+-------+
@end example


@node Selecting columns, Sorting rows, Selecting rows, Retrieving data
@subsubsection Selecting Particular Columns

@cindex columns, selecting
@cindex tables, selecting columns

If you don't want to see entire rows from your table, just name the columns
in which you're interested, separated by commas.  For example, if you want to
know when your animals were born, select the @code{name} and @code{birth}
columns:

@example
mysql> SELECT name, birth FROM pet;
+----------+------------+
| name     | birth      |
+----------+------------+
| Fluffy   | 1993-02-04 |
| Claws    | 1994-03-17 |
| Buffy    | 1989-05-13 |
| Fang     | 1990-08-27 |
| Bowser   | 1989-08-31 |
| Chirpy   | 1998-09-11 |
| Whistler | 1997-12-09 |
| Slim     | 1996-04-29 |
| Puffball | 1999-03-30 |
+----------+------------+
@end example

To find out who owns pets, use this query:

@example
mysql> SELECT owner FROM pet;
+--------+
| owner  |
+--------+
| Harold |
| Gwen   |
| Harold |
| Benny  |
| Diane  |
| Gwen   |
| Gwen   |
| Benny  |
| Diane  |
+--------+
@end example

@findex DISTINCT
However, notice that the query simply retrieves the @code{owner} field from
each record, and some of them appear more than once.  To minimise the output,
retrieve each unique output record just once by adding the keyword
@code{DISTINCT}:

@example
mysql> SELECT DISTINCT owner FROM pet;
+--------+
| owner  |
+--------+
| Benny  |
| Diane  |
| Gwen   |
| Harold |
+--------+
@end example

You can use a @code{WHERE} clause to combine row selection with column
selection.  For example, to get birth dates for dogs and cats only,
use this query:

@example
mysql> SELECT name, species, birth FROM pet
    -> WHERE species = "dog" OR species = "cat";
+--------+---------+------------+
| name   | species | birth      |
+--------+---------+------------+
| Fluffy | cat     | 1993-02-04 |
| Claws  | cat     | 1994-03-17 |
| Buffy  | dog     | 1989-05-13 |
| Fang   | dog     | 1990-08-27 |
| Bowser | dog     | 1989-08-31 |
+--------+---------+------------+
@end example


@node Sorting rows, Date calculations, Selecting columns, Retrieving data
@subsubsection Sorting Rows

@cindex rows, sorting
@cindex sorting, table rows
@cindex sorting, data
@cindex tables, sorting rows

You may have noticed in the preceding examples that the result rows are
displayed in no particular order.  However, it's often easier to examine
query output when the rows are sorted in some meaningful way.  To sort a
result, use an @code{ORDER BY} clause.

Here are animal birthdays, sorted by date:

@example
mysql> SELECT name, birth FROM pet ORDER BY birth;
+----------+------------+
| name     | birth      |
+----------+------------+
| Buffy    | 1989-05-13 |
| Bowser   | 1989-08-31 |
| Fang     | 1990-08-27 |
| Fluffy   | 1993-02-04 |
| Claws    | 1994-03-17 |
| Slim     | 1996-04-29 |
| Whistler | 1997-12-09 |
| Chirpy   | 1998-09-11 |
| Puffball | 1999-03-30 |
+----------+------------+
@end example

On character type columns, sorting@-like all other comparison
operations@-is normally performed in a case-insensitive fashion.
This means that the order will be undefined for columns that are identical
except for their case. You can force a case-sensitive sort by using the
BINARY cast: @code{ORDER BY BINARY(field)}.

To sort in reverse order, add the @code{DESC} (descending) keyword to the
name of the column you are sorting by:

@example
mysql> SELECT name, birth FROM pet ORDER BY birth DESC;
+----------+------------+
| name     | birth      |
+----------+------------+
| Puffball | 1999-03-30 |
| Chirpy   | 1998-09-11 |
| Whistler | 1997-12-09 |
| Slim     | 1996-04-29 |
| Claws    | 1994-03-17 |
| Fluffy   | 1993-02-04 |
| Fang     | 1990-08-27 |
| Bowser   | 1989-08-31 |
| Buffy    | 1989-05-13 |
+----------+------------+
@end example

You can sort on multiple columns.  For example, to sort by type of
animal, then by birth date within animal type with youngest animals first,
use the following query:

@example
mysql> SELECT name, species, birth FROM pet ORDER BY species, birth DESC;
+----------+---------+------------+
| name     | species | birth      |
+----------+---------+------------+
| Chirpy   | bird    | 1998-09-11 |
| Whistler | bird    | 1997-12-09 |
| Claws    | cat     | 1994-03-17 |
| Fluffy   | cat     | 1993-02-04 |
| Fang     | dog     | 1990-08-27 |
| Bowser   | dog     | 1989-08-31 |
| Buffy    | dog     | 1989-05-13 |
| Puffball | hamster | 1999-03-30 |
| Slim     | snake   | 1996-04-29 |
+----------+---------+------------+
@end example

Note that the @code{DESC} keyword applies only to the column name immediately
preceding it (@code{birth}); @code{species} values are still sorted in
ascending order.


@node Date calculations, Working with NULL, Sorting rows, Retrieving data
@subsubsection Date Calculations

@cindex date calculations
@cindex calculating, dates
@cindex extracting, dates
@cindex age, calculating

MySQL provides several functions that you can use to perform
calculations on dates, for example, to calculate ages or extract
parts of dates.

To determine how many years old each of your pets is, compute the
difference in the year part of the current date and the birth date, then
subtract one if the current date occurs earlier in the calendar year than
the birth date.  The following query shows, for each pet, the birth date,
the current date, and the age in years.

@example
mysql> SELECT name, birth, CURRENT_DATE,
    -> (YEAR(CURRENT_DATE)-YEAR(birth))
    -> - (RIGHT(CURRENT_DATE,5)<RIGHT(birth,5))
    -> AS age
    -> FROM pet;
+----------+------------+--------------+------+
| name     | birth      | CURRENT_DATE | age  |
+----------+------------+--------------+------+
| Fluffy   | 1993-02-04 | 2001-08-29   |    8 |
| Claws    | 1994-03-17 | 2001-08-29   |    7 |
| Buffy    | 1989-05-13 | 2001-08-29   |   12 |
| Fang     | 1990-08-27 | 2001-08-29   |   11 |
| Bowser   | 1989-08-31 | 2001-08-29   |   11 |
| Chirpy   | 1998-09-11 | 2001-08-29   |    2 |
| Whistler | 1997-12-09 | 2001-08-29   |    3 |
| Slim     | 1996-04-29 | 2001-08-29   |    5 |
| Puffball | 1999-03-30 | 2001-08-29   |    2 |
+----------+------------+--------------+------+
@end example

Here, @code{YEAR()} pulls out the year part of a date and @code{RIGHT()}
pulls off the rightmost five characters that represent the @code{MM-DD}
(calendar year) part of the date.  The part of the expression that
compares the @code{MM-DD} values evaluates to 1 or 0, which adjusts the
year difference down a year if @code{CURRENT_DATE} occurs earlier in
the year than @code{birth}.  The full expression is somewhat ungainly,
so an alias (@code{age}) is used to make the output column label more
meaningful.

The query works, but the result could be scanned more easily if the rows
were presented in some order.  This can be done by adding an @code{ORDER
BY name} clause to sort the output by name:

@example
mysql> SELECT name, birth, CURRENT_DATE,
    -> (YEAR(CURRENT_DATE)-YEAR(birth))
    -> - (RIGHT(CURRENT_DATE,5)<RIGHT(birth,5))
    -> AS age
    -> FROM pet ORDER BY name;
+----------+------------+--------------+------+
| name     | birth      | CURRENT_DATE | age  |
+----------+------------+--------------+------+
| Bowser   | 1989-08-31 | 2001-08-29   |   11 |
| Buffy    | 1989-05-13 | 2001-08-29   |   12 |
| Chirpy   | 1998-09-11 | 2001-08-29   |    2 |
| Claws    | 1994-03-17 | 2001-08-29   |    7 |
| Fang     | 1990-08-27 | 2001-08-29   |   11 |
| Fluffy   | 1993-02-04 | 2001-08-29   |    8 |
| Puffball | 1999-03-30 | 2001-08-29   |    2 |
| Slim     | 1996-04-29 | 2001-08-29   |    5 |
| Whistler | 1997-12-09 | 2001-08-29   |    3 |
+----------+------------+--------------+------+
@end example

To sort the output by @code{age} rather than @code{name}, just use a
different @code{ORDER BY} clause:

@example
mysql> SELECT name, birth, CURRENT_DATE,
    -> (YEAR(CURRENT_DATE)-YEAR(birth))
    -> - (RIGHT(CURRENT_DATE,5)<RIGHT(birth,5))
    -> AS age
    -> FROM pet ORDER BY age;
+----------+------------+--------------+------+
| name     | birth      | CURRENT_DATE | age  |
+----------+------------+--------------+------+
| Chirpy   | 1998-09-11 | 2001-08-29   |    2 |
| Puffball | 1999-03-30 | 2001-08-29   |    2 |
| Whistler | 1997-12-09 | 2001-08-29   |    3 |
| Slim     | 1996-04-29 | 2001-08-29   |    5 |
| Claws    | 1994-03-17 | 2001-08-29   |    7 |
| Fluffy   | 1993-02-04 | 2001-08-29   |    8 |
| Fang     | 1990-08-27 | 2001-08-29   |   11 |
| Bowser   | 1989-08-31 | 2001-08-29   |   11 |
| Buffy    | 1989-05-13 | 2001-08-29   |   12 |
+----------+------------+--------------+------+
@end example

A similar query can be used to determine age at death for animals that have
died.  You determine which animals these are by checking whether the
@code{death} value is @code{NULL}.  Then, for those with non-@code{NULL}
values, compute the difference between the @code{death} and @code{birth}
values:

@example
mysql> SELECT name, birth, death,
    -> (YEAR(death)-YEAR(birth)) - (RIGHT(death,5)<RIGHT(birth,5))
    -> AS age
    -> FROM pet WHERE death IS NOT NULL ORDER BY age;
+--------+------------+------------+------+
| name   | birth      | death      | age  |
+--------+------------+------------+------+
| Bowser | 1989-08-31 | 1995-07-29 |    5 |
+--------+------------+------------+------+
@end example

The query uses @code{death IS NOT NULL} rather than @code{death <> NULL}
because @code{NULL} is a special value.  This is explained later.
@xref{Working with NULL,  , Working with @code{NULL}}.

What if you want to know which animals have birthdays next month?  For this
type of calculation, year and day are irrelevant; you simply want to extract
the month part of the @code{birth} column.  MySQL provides several
date-part extraction functions, such as @code{YEAR()}, @code{MONTH()}, and
@code{DAYOFMONTH()}.  @code{MONTH()} is the appropriate function here.  To
see how it works, run a simple query that displays the value of both
@code{birth} and @code{MONTH(birth)}:

@example
mysql> SELECT name, birth, MONTH(birth) FROM pet;
+----------+------------+--------------+
| name     | birth      | MONTH(birth) |
+----------+------------+--------------+
| Fluffy   | 1993-02-04 |            2 |
| Claws    | 1994-03-17 |            3 |
| Buffy    | 1989-05-13 |            5 |
| Fang     | 1990-08-27 |            8 |
| Bowser   | 1989-08-31 |            8 |
| Chirpy   | 1998-09-11 |            9 |
| Whistler | 1997-12-09 |           12 |
| Slim     | 1996-04-29 |            4 |
| Puffball | 1999-03-30 |            3 |
+----------+------------+--------------+
@end example

Finding animals with birthdays in the upcoming month is easy, too.  Suppose
the current month is April.  Then the month value is @code{4} and you look
for animals born in May (month 5) like this:

@example
mysql> SELECT name, birth FROM pet WHERE MONTH(birth) = 5;
+-------+------------+
| name  | birth      |
+-------+------------+
| Buffy | 1989-05-13 |
+-------+------------+
@end example

There is a small complication if the current month is December, of course.
You don't just add one to the month number (@code{12}) and look for animals
born in month 13, because there is no such month.  Instead, you look for
animals born in January (month 1).

You can even write the query so that it works no matter what the current
month is.  That way you don't have to use a particular month number
in the query.  @code{DATE_ADD()} allows you to add a time interval to a
given date.  If you add a month to the value of @code{NOW()}, then extract
the month part with @code{MONTH()}, the result produces the month in which to
look for birthdays:

@example
mysql> SELECT name, birth FROM pet
    -> WHERE MONTH(birth) = MONTH(DATE_ADD(NOW(), INTERVAL 1 MONTH));
@end example

A different way to accomplish the same task is to add @code{1} to get the
next month after the current one (after using the modulo function (@code{MOD})
to wrap around the month value to @code{0} if it is currently
@code{12}):

@example
mysql> SELECT name, birth FROM pet
    -> WHERE MONTH(birth) = MOD(MONTH(NOW()), 12) + 1;
@end example

Note that @code{MONTH} returns a number between 1 and 12. And
@code{MOD(something,12)} returns a number between 0 and 11. So the
addition has to be after the @code{MOD()}, otherwise we would go from
November (11) to January (1).


@node Working with NULL, Pattern matching, Date calculations, Retrieving data
@subsubsection Working with @code{NULL} Values

@findex NULL
@cindex NULL value

The @code{NULL} value can be surprising until you get used to it.
Conceptually, @code{NULL} means missing value or unknown value and it
is treated somewhat differently than other values.  To test for @code{NULL},
you cannot use the arithmetic comparison operators such as @code{=}, @code{<},
or @code{<>}.  To demonstrate this for yourself, try the following query:

@example
mysql> SELECT 1 = NULL, 1 <> NULL, 1 < NULL, 1 > NULL;
+----------+-----------+----------+----------+
| 1 = NULL | 1 <> NULL | 1 < NULL | 1 > NULL |
+----------+-----------+----------+----------+
|     NULL |      NULL |     NULL |     NULL |
+----------+-----------+----------+----------+
@end example

Clearly you get no meaningful results from these comparisons.  Use
the @code{IS NULL} and @code{IS NOT NULL} operators instead:

@example
mysql> SELECT 1 IS NULL, 1 IS NOT NULL;
+-----------+---------------+
| 1 IS NULL | 1 IS NOT NULL |
+-----------+---------------+
|         0 |             1 |
+-----------+---------------+
@end example

Note that two @code{NULL} are compared as equal is when you do an
@code{GROUP BY}.

In MySQL, 0 or @code{NULL} means false and anything else means true.
The default truth value from a boolean operation is 1.

When doing an @code{ORDER BY}, @code{NULL} values are always sorted first,
even if you are using @code{DESC}.

This special treatment of @code{NULL} is why, in the previous section, it
was necessary to determine which animals are no longer alive using
@code{death IS NOT NULL} instead of @code{death <> NULL}.


@node Pattern matching, Counting rows, Working with NULL, Retrieving data
@subsubsection Pattern Matching

@cindex pattern matching
@cindex matching, patterns
@cindex expressions, extended

MySQL provides standard SQL pattern matching as well as a form of
pattern matching based on extended regular expressions similar to those used
by Unix utilities such as @code{vi}, @code{grep}, and @code{sed}.

SQL pattern matching allows you to use @samp{_} to match any single
character and @samp{%} to match an arbitrary number of characters (including
zero characters).  In MySQL, SQL patterns are case-insensitive by
default.  Some examples are shown here.  Note that you do not use @code{=}
or @code{<>} when you use SQL patterns; use the @code{LIKE} or @code{NOT
LIKE} comparison operators instead.

To find names beginning with @samp{b}:

@example
mysql> SELECT * FROM pet WHERE name LIKE "b%";
+--------+--------+---------+------+------------+------------+
| name   | owner  | species | sex  | birth      | death      |
+--------+--------+---------+------+------------+------------+
| Buffy  | Harold | dog     | f    | 1989-05-13 | NULL       |
| Bowser | Diane  | dog     | m    | 1989-08-31 | 1995-07-29 |
+--------+--------+---------+------+------------+------------+
@end example

To find names ending with @samp{fy}:

@example
mysql> SELECT * FROM pet WHERE name LIKE "%fy";
+--------+--------+---------+------+------------+-------+
| name   | owner  | species | sex  | birth      | death |
+--------+--------+---------+------+------------+-------+
| Fluffy | Harold | cat     | f    | 1993-02-04 | NULL  |
| Buffy  | Harold | dog     | f    | 1989-05-13 | NULL  |
+--------+--------+---------+------+------------+-------+
@end example

To find names containing a @samp{w}:

@example
mysql> SELECT * FROM pet WHERE name LIKE "%w%";
+----------+-------+---------+------+------------+------------+
| name     | owner | species | sex  | birth      | death      |
+----------+-------+---------+------+------------+------------+
| Claws    | Gwen  | cat     | m    | 1994-03-17 | NULL       |
| Bowser   | Diane | dog     | m    | 1989-08-31 | 1995-07-29 |
| Whistler | Gwen  | bird    | NULL | 1997-12-09 | NULL       |
+----------+-------+---------+------+------------+------------+
@end example

To find names containing exactly five characters, use the @samp{_} pattern
character:

@example
mysql> SELECT * FROM pet WHERE name LIKE "_____";
+-------+--------+---------+------+------------+-------+
| name  | owner  | species | sex  | birth      | death |
+-------+--------+---------+------+------------+-------+
| Claws | Gwen   | cat     | m    | 1994-03-17 | NULL  |
| Buffy | Harold | dog     | f    | 1989-05-13 | NULL  |
+-------+--------+---------+------+------------+-------+
@end example

The other type of pattern matching provided by MySQL uses extended
regular expressions.  When you test for a match for this type of pattern, use
the @code{REGEXP} and @code{NOT REGEXP} operators (or @code{RLIKE} and
@code{NOT RLIKE}, which are synonyms).

Some characteristics of extended regular expressions are:

@itemize @bullet
@item
@samp{.} matches any single character.

@item
A character class @samp{[...]} matches any character within the brackets.
For example, @samp{[abc]} matches @samp{a}, @samp{b}, or @samp{c}.  To name a
range of characters, use a dash.  @samp{[a-z]} matches any lowercase letter,
whereas @samp{[0-9]} matches any digit.

@item
@samp{*} matches zero or more instances of the thing preceding it.  For
example, @samp{x*} matches any number of @samp{x} characters,
@samp{[0-9]*} matches any number of digits, and @samp{.*} matches any
number of anything.

@item
The pattern matches if it occurs anywhere in the value being tested.
(SQL patterns match only if they match the entire value.)

@item
To anchor a pattern so that it must match the beginning or end of the value
being tested, use @samp{^} at the beginning or @samp{$} at the end of the
pattern.
@end itemize

To demonstrate how extended regular expressions work, the @code{LIKE} queries
shown previously are rewritten here to use @code{REGEXP}.

To find names beginning with @samp{b}, use @samp{^} to match the beginning of
the name:

@example
mysql> SELECT * FROM pet WHERE name REGEXP "^b";
+--------+--------+---------+------+------------+------------+
| name   | owner  | species | sex  | birth      | death      |
+--------+--------+---------+------+------------+------------+
| Buffy  | Harold | dog     | f    | 1989-05-13 | NULL       |
| Bowser | Diane  | dog     | m    | 1989-08-31 | 1995-07-29 |
+--------+--------+---------+------+------------+------------+
@end example

Prior to MySQL Version 3.23.4, @code{REGEXP} is case-sensitive,
and the previous query will return no rows. To match either lowercase or
uppercase @samp{b}, use this query instead:

@example
mysql> SELECT * FROM pet WHERE name REGEXP "^[bB]";
@end example

From MySQL 3.23.4 on, to force a @code{REGEXP} comparison to
be case-sensitive, use the @code{BINARY} keyword to make one of the
strings a binary string. This query will match only lowercase @samp{b}
at the beginning of a name:

@example
mysql> SELECT * FROM pet WHERE name REGEXP BINARY "^b";
@end example

To find names ending with @samp{fy}, use @samp{$} to match the end of the
name:

@example
mysql> SELECT * FROM pet WHERE name REGEXP "fy$";
+--------+--------+---------+------+------------+-------+
| name   | owner  | species | sex  | birth      | death |
+--------+--------+---------+------+------------+-------+
| Fluffy | Harold | cat     | f    | 1993-02-04 | NULL  |
| Buffy  | Harold | dog     | f    | 1989-05-13 | NULL  |
+--------+--------+---------+------+------------+-------+
@end example

To find names containing a lowercase or uppercase @samp{w}, use this query:

@example
mysql> SELECT * FROM pet WHERE name REGEXP "w";
+----------+-------+---------+------+------------+------------+
| name     | owner | species | sex  | birth      | death      |
+----------+-------+---------+------+------------+------------+
| Claws    | Gwen  | cat     | m    | 1994-03-17 | NULL       |
| Bowser   | Diane | dog     | m    | 1989-08-31 | 1995-07-29 |
| Whistler | Gwen  | bird    | NULL | 1997-12-09 | NULL       |
+----------+-------+---------+------+------------+------------+
@end example

Because a regular expression pattern matches if it occurs anywhere in the
value, it is not necessary in the previous query to put a wildcard on either
side of the pattern to get it to match the entire value like it would be if
you used a SQL pattern.

To find names containing exactly five characters, use @samp{^} and @samp{$}
to match the beginning and end of the name, and five instances of @samp{.}
in between:

@example
mysql> SELECT * FROM pet WHERE name REGEXP "^.....$";
+-------+--------+---------+------+------------+-------+
| name  | owner  | species | sex  | birth      | death |
+-------+--------+---------+------+------------+-------+
| Claws | Gwen   | cat     | m    | 1994-03-17 | NULL  |
| Buffy | Harold | dog     | f    | 1989-05-13 | NULL  |
+-------+--------+---------+------+------------+-------+
@end example

You could also write the previous query using the @samp{@{n@}}
``repeat-@code{n}-times'' operator:

@example
mysql> SELECT * FROM pet WHERE name REGEXP "^.@{5@}$";
+-------+--------+---------+------+------------+-------+
| name  | owner  | species | sex  | birth      | death |
+-------+--------+---------+------+------------+-------+
| Claws | Gwen   | cat     | m    | 1994-03-17 | NULL  |
| Buffy | Harold | dog     | f    | 1989-05-13 | NULL  |
+-------+--------+---------+------+------------+-------+
@end example


@node Counting rows, Multiple tables, Pattern matching, Retrieving data
@subsubsection Counting Rows

@cindex rows, counting
@cindex tables, counting rows
@cindex counting, table rows

Databases are often used to answer the question, ``How often does a certain
type of data occur in a table?''  For example, you might want to know how
many pets you have, or how many pets each owner has, or you might want to
perform various kinds of censuses on your animals.

Counting the total number of animals you have is the same question as ``How
many rows are in the @code{pet} table?'' because there is one record per pet.
The @code{COUNT()} function counts the number of non-@code{NULL} results, so
the query to count your animals looks like this:

@example
mysql> SELECT COUNT(*) FROM pet;
+----------+
| COUNT(*) |
+----------+
|        9 |
+----------+
@end example

Earlier, you retrieved the names of the people who owned pets.  You can
use @code{COUNT()} if you want to find out how many pets each owner has:

@example
mysql> SELECT owner, COUNT(*) FROM pet GROUP BY owner;
+--------+----------+
| owner  | COUNT(*) |
+--------+----------+
| Benny  |        2 |
| Diane  |        2 |
| Gwen   |        3 |
| Harold |        2 |
+--------+----------+
@end example

Note the use of @code{GROUP BY} to group together all records for each
@code{owner}.  Without it, all you get is an error message:

@example
mysql> SELECT owner, COUNT(owner) FROM pet;
ERROR 1140 at line 1: Mixing of GROUP columns (MIN(),MAX(),COUNT()...)
with no GROUP columns is illegal if there is no GROUP BY clause
@end example

@code{COUNT()} and @code{GROUP BY} are useful for characterising your
data in various ways.  The following examples show different ways to
perform animal census operations.

Number of animals per species:

@example
mysql> SELECT species, COUNT(*) FROM pet GROUP BY species;
+---------+----------+
| species | COUNT(*) |
+---------+----------+
| bird    |        2 |
| cat     |        2 |
| dog     |        3 |
| hamster |        1 |
| snake   |        1 |
+---------+----------+
@end example

Number of animals per sex:

@example
mysql> SELECT sex, COUNT(*) FROM pet GROUP BY sex;
+------+----------+
| sex  | COUNT(*) |
+------+----------+
| NULL |        1 |
| f    |        4 |
| m    |        4 |
+------+----------+
@end example

(In this output, @code{NULL} indicates sex unknown.)

Number of animals per combination of species and sex:

@example
mysql> SELECT species, sex, COUNT(*) FROM pet GROUP BY species, sex;
+---------+------+----------+
| species | sex  | COUNT(*) |
+---------+------+----------+
| bird    | NULL |        1 |
| bird    | f    |        1 |
| cat     | f    |        1 |
| cat     | m    |        1 |
| dog     | f    |        1 |
| dog     | m    |        2 |
| hamster | f    |        1 |
| snake   | m    |        1 |
+---------+------+----------+
@end example

You need not retrieve an entire table when you use @code{COUNT()}.  For
example, the previous query, when performed just on dogs and cats, looks like
this:

@example
mysql> SELECT species, sex, COUNT(*) FROM pet
    -> WHERE species = "dog" OR species = "cat"
    -> GROUP BY species, sex;
+---------+------+----------+
| species | sex  | COUNT(*) |
+---------+------+----------+
| cat     | f    |        1 |
| cat     | m    |        1 |
| dog     | f    |        1 |
| dog     | m    |        2 |
+---------+------+----------+
@end example

Or, if you wanted the number of animals per sex only for known-sex animals:

@example
mysql> SELECT species, sex, COUNT(*) FROM pet
    -> WHERE sex IS NOT NULL
    -> GROUP BY species, sex;
+---------+------+----------+
| species | sex  | COUNT(*) |
+---------+------+----------+
| bird    | f    |        1 |
| cat     | f    |        1 |
| cat     | m    |        1 |
| dog     | f    |        1 |
| dog     | m    |        2 |
| hamster | f    |        1 |
| snake   | m    |        1 |
+---------+------+----------+
@end example


@node Multiple tables,  , Counting rows, Retrieving data
@subsubsection Using More Than one Table

@cindex tables, multiple

The @code{pet} table keeps track of which pets you have.  If you want to
record other information about them, such as events in their lives like
visits to the vet or when litters are born, you need another table.  What
should this table look like? It needs:

@itemize @bullet
@item
To contain the pet name so you know which animal each event pertains
to.

@item
A date so you know when the event occurred.

@item
A field to describe the event.

@item
An event type field, if you want to be able to categorise events.
@end itemize

Given these considerations, the @code{CREATE TABLE} statement for the
@code{event} table might look like this:

@example
mysql> CREATE TABLE event (name VARCHAR(20), date DATE,
    -> type VARCHAR(15), remark VARCHAR(255));
@end example

As with the @code{pet} table, it's easiest to load the initial records
by creating a tab-delimited text file containing the information:

@multitable @columnfractions .10 .15 .10 .35
@item @strong{name} @tab @strong{date} @tab @strong{type} @tab @strong{remark}
@item Fluffy @tab 1995-05-15 @tab litter @tab 4 kittens, 3 female, 1 male
@item Buffy @tab 1993-06-23 @tab litter @tab 5 puppies, 2 female, 3 male
@item Buffy @tab 1994-06-19 @tab litter @tab 3 puppies, 3 female
@item Chirpy @tab 1999-03-21 @tab vet @tab needed beak straightened
@item Slim @tab 1997-08-03 @tab vet @tab broken rib
@item Bowser @tab 1991-10-12 @tab kennel @tab
@item Fang @tab 1991-10-12 @tab kennel @tab
@item Fang @tab 1998-08-28 @tab birthday @tab Gave him a new chew toy
@item Claws @tab 1998-03-17 @tab birthday @tab Gave him a new flea collar
@item Whistler @tab 1998-12-09 @tab birthday @tab First birthday
@end multitable

Load the records like this:

@example
mysql> LOAD DATA LOCAL INFILE "event.txt" INTO TABLE event;
@end example

Based on what you've learned from the queries you've run on the @code{pet}
table, you should be able to perform retrievals on the records in the
@code{event} table; the principles are the same.  But when is the
@code{event} table by itself insufficient to answer questions you might ask?

Suppose you want to find out the ages of each pet when they had their
litters.  The @code{event} table indicates when this occurred, but to
calculate the age of the mother, you need her birth date.  Because that is
stored in the @code{pet} table, you need both tables for the query:

@example
mysql> SELECT pet.name,
    -> (TO_DAYS(date) - TO_DAYS(birth))/365 AS age,
    -> remark
    -> FROM pet, event
    -> WHERE pet.name = event.name AND type = "litter";
+--------+------+-----------------------------+
| name   | age  | remark                      |
+--------+------+-----------------------------+
| Fluffy | 2.27 | 4 kittens, 3 female, 1 male |
| Buffy  | 4.12 | 5 puppies, 2 female, 3 male |
| Buffy  | 5.10 | 3 puppies, 3 female         |
+--------+------+-----------------------------+
@end example

There are several things to note about this query:

@itemize @bullet
@item
The @code{FROM} clause lists two tables because the query needs to pull
information from both of them.

@item
When combining (joining) information from multiple tables, you need to
specify how records in one table can be matched to records in the other.
This is easy because they both have a @code{name} column.  The query uses
@code{WHERE} clause to match up records in the two tables based on the
@code{name} values.

@item
Because the @code{name} column occurs in both tables, you must be specific
about which table you mean when referring to the column.  This is done
by prepending the table name to the column name.
@end itemize

You need not have two different tables to perform a join.  Sometimes it is
useful to join a table to itself, if you want to compare records in a table
to other records in that same table.  For example, to find breeding pairs
among your pets, you can join the @code{pet} table with itself to pair up
males and females of like species:

@example
mysql> SELECT p1.name, p1.sex, p2.name, p2.sex, p1.species
    -> FROM pet AS p1, pet AS p2
    -> WHERE p1.species = p2.species AND p1.sex = "f" AND p2.sex = "m";
+--------+------+--------+------+---------+
| name   | sex  | name   | sex  | species |
+--------+------+--------+------+---------+
| Fluffy | f    | Claws  | m    | cat     |
| Buffy  | f    | Fang   | m    | dog     |
| Buffy  | f    | Bowser | m    | dog     |
+--------+------+--------+------+---------+
@end example

In this query, we specify aliases for the table name in order
to refer to the columns and keep straight which instance of the table
each column reference is associated with.


@node Getting information, Examples, Database use, Tutorial
@section Getting Information About Databases and Tables

@cindex databases, information about
@cindex tables, information about
@findex DESCRIBE

What if you forget the name of a database or table, or what the structure of
a given table is (for example, what its columns are called)?  MySQL
addresses this problem through several statements that provide information
about the databases and tables it supports.

You have already seen @code{SHOW DATABASES}, which lists the databases
managed by the server.  To find out which database is currently selected,
use the @code{DATABASE()} function:

@example
mysql> SELECT DATABASE();
+------------+
| DATABASE() |
+------------+
| menagerie  |
+------------+
@end example

If you haven't selected any database yet, the result is blank.

To find out what tables the current database contains (for example, when
you're not sure about the name of a table), use this command:

@example
mysql> SHOW TABLES;
+---------------------+
| Tables in menagerie |
+---------------------+
| event               |
| pet                 |
+---------------------+
@end example

If you want to find out about the structure of a table, the @code{DESCRIBE}
command is useful; it displays information about each of a table's columns:

@example
mysql> DESCRIBE pet;
+---------+-------------+------+-----+---------+-------+
| Field   | Type        | Null | Key | Default | Extra |
+---------+-------------+------+-----+---------+-------+
| name    | varchar(20) | YES  |     | NULL    |       |
| owner   | varchar(20) | YES  |     | NULL    |       |
| species | varchar(20) | YES  |     | NULL    |       |
| sex     | char(1)     | YES  |     | NULL    |       |
| birth   | date        | YES  |     | NULL    |       |
| death   | date        | YES  |     | NULL    |       |
+---------+-------------+------+-----+---------+-------+
@end example

@code{Field} indicates the column name, @code{Type} is the data type for
the column, @code{NULL} indicates whether the column can contain
@code{NULL} values, @code{Key} indicates whether the column is
indexed, and @code{Default} specifies the column's default value.

If you have indexes on a table,
@code{SHOW INDEX FROM tbl_name} produces information about them.


@node Examples, Batch mode, Getting information, Tutorial
@section Examples of Common Queries

@cindex queries, examples
@cindex examples, queries

Here are examples of how to solve some common problems with
MySQL.

Some of the examples use the table @code{shop} to hold the price of each
article (item number) for certain traders (dealers).  Supposing that each
trader has a single fixed price per article, then (@code{article},
@code{dealer}) is a primary key for the records.

Start the command-line tool @code{mysql} and select a database:

@example
mysql your-database-name
@end example

(In most MySQL installations, you can use the database-name 'test').

You can create the example table as:

@example
CREATE TABLE shop (
 article INT(4) UNSIGNED ZEROFILL DEFAULT '0000' NOT NULL,
 dealer  CHAR(20)                 DEFAULT ''     NOT NULL,
 price   DOUBLE(16,2)             DEFAULT '0.00' NOT NULL,
 PRIMARY KEY(article, dealer));

INSERT INTO shop VALUES
(1,'A',3.45),(1,'B',3.99),(2,'A',10.99),(3,'B',1.45),(3,'C',1.69),
(3,'D',1.25),(4,'D',19.95);
@end example

Okay, so the example data is:

@example
mysql> SELECT * FROM shop;

+---------+--------+-------+
| article | dealer | price |
+---------+--------+-------+
|    0001 | A      |  3.45 |
|    0001 | B      |  3.99 |
|    0002 | A      | 10.99 |
|    0003 | B      |  1.45 |
|    0003 | C      |  1.69 |
|    0003 | D      |  1.25 |
|    0004 | D      | 19.95 |
+---------+--------+-------+
@end example

@menu
* example-Maximum-column::      The maximum value for a column
* example-Maximum-row::         The row holding the maximum of a certain column
* example-Maximum-column-group::  Maximum of column per group
* example-Maximum-column-group-row::  The rows holding the group-wise maximum of a certain field
* example-user-variables::      Using user variables
* example-Foreign keys::        Using foreign keys
* Searching on two keys::       Searching on Two Keys
* Calculating days::            Calculating Visits Per Day
* example-AUTO_INCREMENT::      Using AUTO_INCREMENT
@end menu


@node example-Maximum-column, example-Maximum-row, Examples, Examples
@subsection The Maximum Value for a Column

``What's the highest item number?''

@example
SELECT MAX(article) AS article FROM shop

+---------+
| article |
+---------+
|       4 |
+---------+
@end example


@node example-Maximum-row, example-Maximum-column-group, example-Maximum-column, Examples
@subsection The Row Holding the Maximum of a Certain Column

``Find number, dealer, and price of the most expensive article.''

In ANSI SQL this is easily done with a sub-query:

@example
SELECT article, dealer, price
FROM   shop
WHERE  price=(SELECT MAX(price) FROM shop)
@end example

In MySQL (which does not yet have sub-selects), just do it in
two steps:

@enumerate
@item
Get the maximum price value from the table with a @code{SELECT} statement.
@item
Using this value compile the actual query:
@example
SELECT article, dealer, price
FROM   shop
WHERE  price=19.95
@end example
@end enumerate

Another solution is to sort all rows descending by price and only
get the first row using the MySQL specific @code{LIMIT} clause:

@example
SELECT article, dealer, price
FROM   shop
ORDER BY price DESC
LIMIT 1
@end example

@strong{NOTE}:  If there are several most expensive articles (for example, each 19.95)
the @code{LIMIT} solution shows only one of them!


@node example-Maximum-column-group, example-Maximum-column-group-row, example-Maximum-row, Examples
@subsection Maximum of Column per Group

``What's the highest price per article?''

@example
SELECT article, MAX(price) AS price
FROM   shop
GROUP BY article

+---------+-------+
| article | price |
+---------+-------+
|    0001 |  3.99 |
|    0002 | 10.99 |
|    0003 |  1.69 |
|    0004 | 19.95 |
+---------+-------+
@end example


@node example-Maximum-column-group-row, example-user-variables, example-Maximum-column-group, Examples
@subsection The Rows Holding the Group-wise Maximum of a Certain Field

``For each article, find the dealer(s) with the most expensive price.''

In ANSI SQL, I'd do it with a sub-query like this:

@example
SELECT article, dealer, price
FROM   shop s1
WHERE  price=(SELECT MAX(s2.price)
              FROM shop s2
              WHERE s1.article = s2.article);
@end example

In MySQL it's best do it in several steps:

@enumerate
@item
Get the list of (article,maxprice).
@item
For each article get the corresponding rows that have the stored maximum
price.
@end enumerate

This can easily be done with a temporary table:

@example
CREATE TEMPORARY TABLE tmp (
        article INT(4) UNSIGNED ZEROFILL DEFAULT '0000' NOT NULL,
        price   DOUBLE(16,2)             DEFAULT '0.00' NOT NULL);

LOCK TABLES shop read;

INSERT INTO tmp SELECT article, MAX(price) FROM shop GROUP BY article;

SELECT shop.article, dealer, shop.price FROM shop, tmp
WHERE shop.article=tmp.article AND shop.price=tmp.price;

UNLOCK TABLES;

DROP TABLE tmp;
@end example

If you don't use a @code{TEMPORARY} table, you must also lock the 'tmp' table.

``Can it be done with a single query?''

Yes, but only by using a quite inefficient trick that I call the
``MAX-CONCAT trick'':

@example
SELECT article,
       SUBSTRING( MAX( CONCAT(LPAD(price,6,'0'),dealer) ), 7) AS dealer,
  0.00+LEFT(      MAX( CONCAT(LPAD(price,6,'0'),dealer) ), 6) AS price
FROM   shop
GROUP BY article;

+---------+--------+-------+
| article | dealer | price |
+---------+--------+-------+
|    0001 | B      |  3.99 |
|    0002 | A      | 10.99 |
|    0003 | C      |  1.69 |
|    0004 | D      | 19.95 |
+---------+--------+-------+
@end example

The last example can, of course, be made a bit more efficient by doing the
splitting of the concatenated column in the client.


@node example-user-variables, example-Foreign keys, example-Maximum-column-group-row, Examples
@subsection Using user variables

You can use MySQL user variables to remember results without
having to store them in temporary variables in the client.
@xref{Variables}.

For example, to find the articles with the highest and lowest price you
can do:

@example
mysql> SELECT @@min_price:=MIN(price),@@max_price:=MAX(price) FROM shop;
mysql> SELECT * FROM shop WHERE price=@@min_price OR price=@@max_price;
+---------+--------+-------+
| article | dealer | price |
+---------+--------+-------+
|    0003 | D      |  1.25 |
|    0004 | D      | 19.95 |
+---------+--------+-------+
@end example


@node example-Foreign keys, Searching on two keys, example-user-variables, Examples
@subsection Using Foreign Keys

@cindex foreign keys
@cindex keys, foreign

In MySQL 3.23.44 and up, @code{InnoDB} tables supports checking of
foreign key constraints. @xref{InnoDB}.
See also @ref{ANSI diff Foreign Keys}.

You don't actually need foreign keys to join 2 tables.
The only thing MySQL currently doesn't do (in type types other than
@code{InnoDB}), is @code{CHECK} to make sure that the keys you use
really exist in the table(s) you're referencing and it
doesn't automatically delete rows from a table with a foreign key
definition. If you use your keys like normal, it'll work just fine:


@example
CREATE TABLE person (
    id SMALLINT UNSIGNED NOT NULL AUTO_INCREMENT,
    name CHAR(60) NOT NULL,
    PRIMARY KEY (id)
);

CREATE TABLE shirt (
    id SMALLINT UNSIGNED NOT NULL AUTO_INCREMENT,
    style ENUM('t-shirt', 'polo', 'dress') NOT NULL,
    color ENUM('red', 'blue', 'orange', 'white', 'black') NOT NULL,
    owner SMALLINT UNSIGNED NOT NULL REFERENCES persons,
    PRIMARY KEY (id)
);


INSERT INTO person VALUES (NULL, 'Antonio Paz');

INSERT INTO shirt VALUES
(NULL, 'polo', 'blue', LAST_INSERT_ID()),
(NULL, 'dress', 'white', LAST_INSERT_ID()),
(NULL, 't-shirt', 'blue', LAST_INSERT_ID());


INSERT INTO person VALUES (NULL, 'Lilliana Angelovska');

INSERT INTO shirt VALUES
(NULL, 'dress', 'orange', LAST_INSERT_ID()),
(NULL, 'polo', 'red', LAST_INSERT_ID()),
(NULL, 'dress', 'blue', LAST_INSERT_ID()),
(NULL, 't-shirt', 'white', LAST_INSERT_ID());


SELECT * FROM person;
+----+---------------------+
| id | name                |
+----+---------------------+
|  1 | Antonio Paz         |
|  2 | Lilliana Angelovska |
+----+---------------------+

SELECT * FROM shirt;
+----+---------+--------+-------+
| id | style   | color  | owner |
+----+---------+--------+-------+
|  1 | polo    | blue   |     1 |
|  2 | dress   | white  |     1 |
|  3 | t-shirt | blue   |     1 |
|  4 | dress   | orange |     2 |
|  5 | polo    | red    |     2 |
|  6 | dress   | blue   |     2 |
|  7 | t-shirt | white  |     2 |
+----+---------+--------+-------+


SELECT s.* FROM person p, shirt s
 WHERE p.name LIKE 'Lilliana%'
   AND s.owner = p.id
   AND s.color <> 'white';

+----+-------+--------+-------+
| id | style | color  | owner |
+----+-------+--------+-------+
|  4 | dress | orange |     2 |
|  5 | polo  | red    |     2 |
|  6 | dress | blue   |     2 |
+----+-------+--------+-------+
@end example


@node Searching on two keys, Calculating days, example-Foreign keys, Examples
@subsection Searching on Two Keys

@findex UNION
@cindex searching, two keys
@cindex keys, searching on two

MySQL doesn't yet optimise when you search on two different
keys combined with @code{OR} (searching on one key with different @code{OR}
parts is optimised quite well):

@example
SELECT field1_index, field2_index FROM test_table WHERE field1_index = '1'
OR  field2_index = '1'
@end example

The reason is that we haven't yet had time to come up with an efficient
way to handle this in the general case. (The @code{AND} handling is,
in comparison, now completely general and works very well.)

For the moment you can solve this very efficiently by using a
@code{TEMPORARY} table. This type of optimisation is also very good if
you are using very complicated queries where the SQL server does the
optimisations in the wrong order.

@example
CREATE TEMPORARY TABLE tmp
SELECT field1_index, field2_index FROM test_table WHERE field1_index = '1';
INSERT INTO tmp
SELECT field1_index, field2_index FROM test_table WHERE field2_index = '1';
SELECT * from tmp;
DROP TABLE tmp;
@end example

The above way to solve this query is in effect a @code{UNION} of two queries.
@xref{UNION}.

@node Calculating days, example-AUTO_INCREMENT, Searching on two keys, Examples
@subsection Calculating Visits Per Day

@findex BIT_OR
@findex BIT_COUNT
@findex <<
@cindex bit_functions, example

The following shows an idea of how you can use the bit group functions
to calculate the number of days per month a user has visited a web page.

@example
CREATE TABLE t1 (year YEAR(4), month INT(2) UNSIGNED ZEROFILL,
             day INT(2) UNSIGNED ZEROFILL);
INSERT INTO t1 VALUES(2000,1,1),(2000,1,20),(2000,1,30),(2000,2,2),
            (2000,2,23),(2000,2,23);
SELECT year,month,BIT_COUNT(BIT_OR(1<<day)) AS days FROM t1
       GROUP BY year,month;

Which returns:

+------+-------+------+
| year | month | days |
+------+-------+------+
| 2000 |    01 |    3 |
| 2000 |    02 |    2 |
+------+-------+------+
@end example

The above calculates how many different days was used for a given
year/month combination, with automatic removal of duplicate entries.


@node example-AUTO_INCREMENT,  , Calculating days, Examples
@subsection Using @code{AUTO_INCREMENT}

@cindex AUTO_INCREMENT
The @code{AUTO_INCREMENT} attribute can be used to generate an unique
identity for new rows:

@example
CREATE TABLE animals (id MEDIUMINT NOT NULL AUTO_INCREMENT,
name CHAR(30) NOT NULL, PRIMARY KEY (id));
INSERT INTO animals (name) VALUES ("dog"),("cat"),("penguin"),
("lax"),("whale");
SELECT * FROM animals;

Which returns:

+----+---------+
| id | name    |
+----+---------+
|  1 | dog     |
|  2 | cat     |
|  3 | penguin |
|  4 | lax     |
|  5 | whale   |
+----+---------+
@end example

For MyISAM and BDB tables you can specify @code{AUTO_INCREMENT} on
secondary column in a multi-column key.  In this case the generated
value for the autoincrement column is calculated as
@code{MAX(auto_increment_column)+1) WHERE prefix=given-prefix}.  This is
useful when you want to put data into ordered groups.

@example
CREATE TABLE animals (grp ENUM('fish','mammal','bird') NOT NULL,
             id MEDIUMINT NOT NULL AUTO_INCREMENT
             PRIMARY KEY (grp,id));
INSERT INTO animals (grp,name) VALUES("mammal","dog"),("mammal","cat"),
            ("bird","penguin"),("fish","lax"),("mammal","whale");
SELECT * FROM animals ORDER BY grp,id;

Which returns:

+--------+----+---------+
| grp    | id | name    |
+--------+----+---------+
| fish   |  1 | lax     |
| mammal |  1 | dog     |
| mammal |  2 | cat     |
| mammal |  3 | whale   |
| bird   |  1 | penguin |
+--------+----+---------+
@end example

Note that in this case, the @code{AUTO_INCREMENT} value will be reused if you
delete the row with the biggest @code{AUTO_INCREMENT} value in any group.

You can get the used @code{AUTO_INCREMENT} key with the
@code{LAST_INSERT_ID()} SQL function or the @code{mysql_insert_id()} API
function.

@node Batch mode, Twin, Examples, Tutorial
@section Using @code{mysql} in Batch Mode

@cindex modes, batch
@cindex batch mode
@cindex running, batch mode
@cindex script files
@cindex files, script

In the previous sections, you used @code{mysql} interactively to enter
queries and view the results.  You can also run @code{mysql} in batch
mode.  To do this, put the commands you want to run in a file, then
tell @code{mysql} to read its input from the file:

@example
shell> mysql < batch-file
@end example

If you are running @code{mysql} under windows and have some special
characters in the file that causes problems, you can do:

@example
dos> mysql -e "source batch-file"
@end example

If you need to specify connection parameters on the command-line, the
command might look like this:

@example
shell> mysql -h host -u user -p < batch-file
Enter password: ********
@end example

When you use @code{mysql} this way, you are creating a script file, then
executing the script.

If you want the script to continue even if you have errors, you should
use the @code{--force} command-line option.

Why use a script?  Here are a few reasons:

@itemize @bullet
@item
If you run a query repeatedly (say, every day or every week), making it a
script allows you to avoid retyping it each time you execute it.

@item
You can generate new queries from existing ones that are similar by copying
and editing script files.

@item
Batch mode can also be useful while you're developing a query, particularly
for multiple-line commands or multiple-statement sequences of commands.  If
you make a mistake, you don't have to retype everything.  Just edit your
script to correct the error, then tell @code{mysql} to execute it again.

@item
If you have a query that produces a lot of output, you can run the output
through a pager rather than watching it scroll off the top of your screen:

@example
shell> mysql < batch-file | more
@end example

@item
You can catch the output in a file for further processing:

@example
shell> mysql < batch-file > mysql.out
@end example

@item
You can distribute your script to other people so they can run the commands,
too.

@item
Some situations do not allow for interactive use, for example, when you run
a query from a @code{cron} job.  In this case, you must use batch mode.
@end itemize

The default output format is different (more concise) when you run
@code{mysql} in batch mode than when you use it interactively.  For
example, the output of @code{SELECT DISTINCT species FROM pet} looks like
this when run interactively:

@example
+---------+
| species |
+---------+
| bird    |
| cat     |
| dog     |
| hamster |
| snake   |
+---------+
@end example

But like this when run in batch mode:

@example
species
bird
cat
dog
hamster
snake
@end example

If you want to get the interactive output format in batch mode, use
@code{mysql -t}.  To echo to the output the commands that are executed, use
@code{mysql -vvv}.

You can also use scripts in the @code{mysql} command-line prompt by
using the @code{source} command:

@example
mysql> source filename;
@end example

@node Twin, Apache, Batch mode, Tutorial
@section Queries from Twin Project

@cindex Twin Studies, queries
@cindex queries, Twin Studeis project

At Analytikerna and Lentus, we have been doing the systems and field work
for a big research project. This project is a collaboration between the
Institute of Environmental Medicine at Karolinska Institutet Stockholm
and the Section on Clinical Research in Aging and Psychology at the
University of Southern California.

The project involves a screening part where all twins in Sweden older
than 65 years are interviewed by telephone. Twins who meet certain
criteria are passed on to the next stage. In this latter stage, twins who
want to participate are visited by a doctor/nurse team. Some of the
examinations include physical and neuropsychological examination,
laboratory testing, neuroimaging, psychological status assessment, and family
history collection. In addition, data are collected on medical and
environmental risk factors.

More information about Twin studies can be found at:
@uref{http://www.imm.ki.se/TWIN/TWINUKW.HTM}

The latter part of the project is administered with a web interface
written using Perl and MySQL.

Each night all data from the interviews are moved into a MySQL
database.

@menu
* Twin pool::                   Find all non-distributed twins
* Twin event::                  Show a table on twin pair status
@end menu


@node Twin pool, Twin event, Twin, Twin
@subsection Find all Non-distributed Twins

The following query is used to determine who goes into the second part of the
project:

@example
SELECT
        CONCAT(p1.id, p1.tvab) + 0 AS tvid,
        CONCAT(p1.christian_name, " ", p1.surname) AS Name,
        p1.postal_code AS Code,
        p1.city AS City,
        pg.abrev AS Area,
        IF(td.participation = "Aborted", "A", " ") AS A,
        p1.dead AS dead1,
        l.event AS event1,
        td.suspect AS tsuspect1,
        id.suspect AS isuspect1,
        td.severe AS tsevere1,
        id.severe AS isevere1,
        p2.dead AS dead2,
        l2.event AS event2,
        h2.nurse AS nurse2,
        h2.doctor AS doctor2,
        td2.suspect AS tsuspect2,
        id2.suspect AS isuspect2,
        td2.severe AS tsevere2,
        id2.severe AS isevere2,
        l.finish_date
FROM
        twin_project AS tp
        /* For Twin 1 */
        LEFT JOIN twin_data AS td ON tp.id = td.id
                  AND tp.tvab = td.tvab
        LEFT JOIN informant_data AS id ON tp.id = id.id
                  AND tp.tvab = id.tvab
        LEFT JOIN harmony AS h ON tp.id = h.id
                  AND tp.tvab = h.tvab
        LEFT JOIN lentus AS l ON tp.id = l.id
                  AND tp.tvab = l.tvab
        /* For Twin 2 */
        LEFT JOIN twin_data AS td2 ON p2.id = td2.id
                  AND p2.tvab = td2.tvab
        LEFT JOIN informant_data AS id2 ON p2.id = id2.id
                  AND p2.tvab = id2.tvab
        LEFT JOIN harmony AS h2 ON p2.id = h2.id
                  AND p2.tvab = h2.tvab
        LEFT JOIN lentus AS l2 ON p2.id = l2.id
                  AND p2.tvab = l2.tvab,
        person_data AS p1,
        person_data AS p2,
        postal_groups AS pg
WHERE
        /* p1 gets main twin and p2 gets his/her twin. */
        /* ptvab is a field inverted from tvab */
        p1.id = tp.id AND p1.tvab = tp.tvab AND
        p2.id = p1.id AND p2.ptvab = p1.tvab AND
        /* Just the sceening survey */
        tp.survey_no = 5 AND
        /* Skip if partner died before 65 but allow emigration (dead=9) */
        (p2.dead = 0 OR p2.dead = 9 OR
         (p2.dead = 1 AND
          (p2.death_date = 0 OR
           (((TO_DAYS(p2.death_date) - TO_DAYS(p2.birthday)) / 365)
            >= 65))))
        AND
        (
        /* Twin is suspect */
        (td.future_contact = 'Yes' AND td.suspect = 2) OR
        /* Twin is suspect - Informant is Blessed */
        (td.future_contact = 'Yes' AND td.suspect = 1
                                   AND id.suspect = 1) OR
        /* No twin - Informant is Blessed */
        (ISNULL(td.suspect) AND id.suspect = 1
                            AND id.future_contact = 'Yes') OR
        /* Twin broken off - Informant is Blessed */
        (td.participation = 'Aborted'
         AND id.suspect = 1 AND id.future_contact = 'Yes') OR
        /* Twin broken off - No inform - Have partner */
        (td.participation = 'Aborted' AND ISNULL(id.suspect)
                                      AND p2.dead = 0))
        AND
        l.event = 'Finished'
        /* Get at area code */
        AND SUBSTRING(p1.postal_code, 1, 2) = pg.code
        /* Not already distributed */
        AND (h.nurse IS NULL OR h.nurse=00 OR h.doctor=00)
        /* Has not refused or been aborted */
        AND NOT (h.status = 'Refused' OR h.status = 'Aborted'
        OR h.status = 'Died' OR h.status = 'Other')
ORDER BY
        tvid;
@end example

Some explanations:
@table @asis
@item @code{CONCAT(p1.id, p1.tvab) + 0 AS tvid}
We want to sort on the concatenated @code{id} and @code{tvab} in
numerical order. Adding @code{0} to the result causes MySQL to
treat the result as a number.
@item column @code{id}
This identifies a pair of twins. It is a key in all tables.
@item column @code{tvab}
This identifies a twin in a pair. It has a value of @code{1} or @code{2}.
@item column @code{ptvab}
This is an inverse of @code{tvab}. When @code{tvab} is @code{1} this is
@code{2}, and vice versa. It exists to save typing and to make it easier for
MySQL to optimise the query.
@end table

This query demonstrates, among other things, how to do lookups on a
table from the same table with a join (@code{p1} and @code{p2}). In the example, this
is used to check whether a twin's partner died before the age of 65. If so,
the row is not returned.

All of the above exist in all tables with twin-related information. We
have a key on both @code{id,tvab} (all tables), and @code{id,ptvab}
(@code{person_data}) to make queries faster.

On our production machine (A 200MHz UltraSPARC), this query returns
about 150-200 rows and takes less than one second.

The current number of records in the tables used above:
@multitable @columnfractions .25 .10
@item @strong{Table}            @tab @strong{Rows}
@item @code{person_data}        @tab 71074
@item @code{lentus}		@tab 5291
@item @code{twin_project}       @tab 5286
@item @code{twin_data}		@tab 2012
@item @code{informant_data}	@tab 663
@item @code{harmony}		@tab 381
@item @code{postal_groups}	@tab 100
@end multitable


@node Twin event,  , Twin pool, Twin
@subsection Show a Table on Twin Pair Status

Each interview ends with a status code called @code{event}. The query
shown here is used to display a table over all twin pairs combined by
event. This indicates in how many pairs both twins are finished, in how many
pairs one twin is finished and the other refused, and so on.

@example
SELECT
        t1.event,
        t2.event,
        COUNT(*)
FROM
        lentus AS t1,
        lentus AS t2,
        twin_project AS tp
WHERE
        /* We are looking at one pair at a time */
        t1.id = tp.id
        AND t1.tvab=tp.tvab
        AND t1.id = t2.id
        /* Just the sceening survey */
        AND tp.survey_no = 5
        /* This makes each pair only appear once */
        AND t1.tvab='1' AND t2.tvab='2'
GROUP BY
        t1.event, t2.event;

@end example


@node Apache,  , Twin, Tutorial
@section Using MySQL with Apache

@cindex Apache

There are programs that let you authenticate your users from a MySQL
database and also let you log your log files into a MySQL table.
@xref{Portals}.

You can change the Apache logging format to be easily readable by
MySQL by putting the following into the Apache configuration file:

@example
LogFormat \
        "\"%h\",%@{%Y%m%d%H%M%S@}t,%>s,\"%b\",\"%@{Content-Type@}o\",  \
        \"%U\",\"%@{Referer@}i\",\"%@{User-Agent@}i\""
@end example

In MySQL you can do something like this:

@example
LOAD DATA INFILE '/local/access_log' INTO TABLE table_name
FIELDS TERMINATED BY ',' OPTIONALLY ENCLOSED BY '"' ESCAPED BY '\\'
@end example




@node MySQL Database Administration, MySQL Optimisation, Tutorial, Top
@chapter Database Administration

@menu
* Configuring MySQL::           Configuring MySQL
* Privilege system::            General Security Issues and the MySQL Access Privilege System
* User Account Management::     MySQL User Account Management
* Disaster Prevention::         Disaster Prevention and Recovery
* Database Administration::     Database Administration Language Reference
* Localisation::                MySQL Localisation and International Usage
* Server-Side Scripts::         MySQL Server-Side Scripts and Utilities
* Client-Side Scripts::         MySQL Client-Side Scripts and Utilities
* Log Files::                   The MySQL Log Files
* Replication::                 Replication in MySQL
@end menu


@node Configuring MySQL, Privilege system, MySQL Database Administration, MySQL Database Administration
@section Configuring MySQL

@c FIX: add section intro


@menu
* Command-line options::        @code{mysqld} Command-line Options
* Option files::                @file{my.cnf} Option Files
* Installing many servers::     Installing Many Servers on the Same Machine
* Multiple servers::            Running Multiple MySQL Servers on the Same Machine
@end menu


@node Command-line options, Option files, Configuring MySQL, Configuring MySQL
@subsection @code{mysqld} Command-line Options

@findex command-line options
@cindex options, command-line
@cindex mysqld options

In most cases you should manage mysqld options through option files.
@xref{Option files}.

@code{mysqld} and @code{mysqld.server} reads options from the
@code{mysqld} and @code{server} groups. @code{mysqld_safe} read options
from the @code{mysqld}, @code{server}, @code{mysqld_safe} and
@code{safe_mysqld} groups.  An embedded MySQL server usually reads
options from the @code{server}, @code{embedded} and
@code{xxxxx_SERVER}, where @code{xxxxx} is the name of the application.


@code{mysqld} accepts the following command-line options:

@table @code
@item --ansi
Use ANSI SQL syntax instead of MySQL syntax.  @xref{ANSI mode}.

@item -b, --basedir=path
Path to installation directory. All paths are usually resolved relative to
this.

@item --big-tables
Allow big result sets by saving all temporary sets on file.  It solves
most 'table full' errors, but also slows down the queries where
in-memory tables would suffice.  Since Version 3.23.2, MySQL is
able to solve it automatically by using memory for small temporary
tables and switching to disk tables where necessary.

@item --bind-address=IP
IP address to bind to.

@item --character-sets-dir=path
Directory where character sets are.  @xref{Character sets}.

@item --chroot=path
Chroot @code{mysqld} daemon during startup.  Recommended security measure. It will
somewhat limit @code{LOAD DATA INFILE} and @code{SELECT ... INTO OUTFILE}
though.

@item --core-file
Write a core file if @code{mysqld} dies.  For some systems you must also
specify @code{--core-file-size} to @code{safe_mysqld}.
@xref{safe_mysqld, ,@code{safe_mysqld}}.
Note that on some systems, like Solaris, you will
not get a core file if you are also using the @code{--user} option.

@item -h, --datadir=path
Path to the database root.

@item --debug[...]=
If MySQL is configured with @code{--with-debug}, you can use this
option to get a trace file of what @code{mysqld} is doing.
@xref{Making trace files}.

@item --default-character-set=charset
Set the default character set.  @xref{Character sets}.

@item --default-table-type=type
Set the default table type for tables.  @xref{Table types}.

@item --delay-key-write-for-all-tables
Don't flush key buffers between writes for any @code{MyISAM} table.
@xref{Server parameters}.

@item --des-key-file=filename
Read the default keys used by @code{DES_ENCRYPT()} and @code{DES_DECRYPT()}
from this file.

@item --enable-locking
Enable system locking.  Note that if you use this option on a system on
which @code{lockd} does not fully work (as on Linux), you will easily get
mysqld to deadlock.

@item --enable-named-pipe
Enable support for named pipes (only on NT/Win2000/XP).

@item -T, --exit-info
This is a bit mask of different flags one can use for debugging the
mysqld server; one should not use this option if one doesn't know
exactly what it does!

@item --flush
Flush all changes to disk after each SQL command.  Normally MySQL
only does a write of all changes to disk after each SQL command and lets
the operating system handle the syncing to disk.
@xref{Crashing}.

@item -?, --help
Display short help and exit.

@item --init-file=file
Read SQL commands from this file at startup.

@item -L, --language=...
Client error messages in given language.  May be given as a full path.
@xref{Languages}.

@item -l, --log[=file]
Log connections and queries to file. @xref{Query log}.

@item --log-isam[=file]
Log all ISAM/MyISAM changes to file (only used when debugging ISAM/MyISAM).

@item --log-slow-queries[=file]
Log all queries that have taken more than @code{long_query_time} seconds to
execute to file.  @xref{Slow query log}.

@item --log-update[=file]
Log updates to @code{file.#} where @code{#} is a unique number if not given.
@xref{Update log}.

@item --log-long-format
Log some extra information to update log.  If you are using
@code{--log-slow-queries} then queries that are not using indexes are logged
to the slow query log.

@item --low-priority-updates
Table-modifying operations (@code{INSERT}/@code{DELETE}/@code{UPDATE})
will have lower priority than selects.  It can also be done via
@code{@{INSERT | REPLACE | UPDATE | DELETE@} LOW_PRIORITY ...} to lower
the priority of only one query, or by
@code{SET OPTION SQL_LOW_PRIORITY_UPDATES=1} to change the priority in one
thread.  @xref{Table locking}.

@item --memlock
Lock the @code{mysqld} process in memory.  This works only if your
system supports the @code{mlockall()} system call (like Solaris).  This
may help if you have a problem where the operating system is causing
@code{mysqld} to swap on disk.

@item --myisam-recover [=option[,option...]]]
Option is any combination
of @code{DEFAULT}, @code{BACKUP}, @code{FORCE} or @code{QUICK}.  You can
also set this explicitly to @code{""} if you want to disable this
option. If this option is used, @code{mysqld} will on open check if the
table is marked as crashed or if the table wasn't closed properly.
(The last option only works if you are running with
@code{--skip-locking}.)  If this is the case @code{mysqld} will run
check on the table. If the table was corrupted, @code{mysqld} will
attempt to repair it.

The following options affects how the repair works.

@multitable @columnfractions .15 .7
@item @strong{Option} @tab @strong{Description}
@item DEFAULT  @tab The same as not giving any option to
                    @code{--myisam-recover}.
@item BACKUP   @tab If the data table was changed during recover, save a
                    backup of the @file{table_name.MYD} datafile as
                    @file{table_name-datetime.BAK}.
@item FORCE    @tab Run recover even if we will loose more than one row
                    from the .MYD file.
@item QUICK    @tab Don't check the rows in the table if there isn't any
                    delete blocks.
@end multitable

Before a table is automatically repaired, MySQL will add a note
about this in the error log.  If you want to be able to recover from most
things without user intervention, you should use the options
@code{BACKUP,FORCE}.  This will force a repair of a table even if some rows
would be deleted, but it will keep the old datafile as a backup so that
you can later examine what happened.

@item --pid-file=path
Path to pid file used by @code{safe_mysqld}.

@item -P, --port=...
Port number to listen for TCP/IP connections.

@item -o, --old-protocol
Use the 3.20 protocol for compatibility with some very old clients.
@xref{Upgrading-from-3.20}.

@item --one-thread
Only use one thread (for debugging under Linux).  @xref{Debugging server}.

@item -O, --set-variable var=option
Give a variable a value. @code{--help} lists variables.  You can find a full
description for all variables in the @code{SHOW VARIABLES} section in this
manual.  @xref{SHOW VARIABLES}.  The tuning server parameters section includes
information of how to optimise these.  @xref{Server parameters}.

@item --safe-mode
Skip some optimise stages.  Implies @code{--skip-delay-key-write}.

@item --safe-show-database
Don't show databases for which the user doesn't have any privileges.

@item --safe-user-create
If this is enabled, a user can't create new users with the GRANT
command, if the user doesn't have @strong{insert} privilege to the
@code{mysql.user} table or any column in this table.

@item --skip-concurrent-insert
Turn off the ability to select and insert at the same time on @code{MyISAM}
tables. (This is only to be used if you think you have found a bug in this
feature.)

@item --skip-delay-key-write
Ignore the @code{delay_key_write} option for all tables.
@xref{Server parameters}.

@item --skip-grant-tables
This option causes the server not to use the privilege system at all.  This
gives everyone @strong{full access} to all databases!  (You can tell a running
server to start using the grant tables again by executing @code{mysqladmin
flush-privileges} or @code{mysqladmin reload}.)

@item --skip-host-cache
Never use host name cache for faster name-ip resolution, but query DNS server
on every connect instead.  @xref{DNS}.

@item --skip-locking
Don't use system locking.  To use @code{isamchk} or @code{myisamchk} you must
shut down the server.  @xref{Stability}.  Note that in MySQL Version
3.23 you can use @code{REPAIR} and @code{CHECK} to repair/check @code{MyISAM}
tables.

@item --skip-name-resolve
Hostnames are not resolved.  All @code{Host} column values in the grant tables
must be IP numbers or @code{localhost}.  @xref{DNS}.

@item --skip-networking
Don't listen for TCP/IP connections at all.  All interaction with
@code{mysqld} must be made via Unix sockets.  This option is highly
recommended for systems where only local requests are allowed.  @xref{DNS}.

@item --skip-new
Don't use new, possible wrong routines.  Implies @code{--skip-delay-key-write}.
This will also set default table type to @code{ISAM}.  @xref{ISAM}.

@item --skip-symlink
Don't delete or rename files that a symlinked file in the data directory
points to.

@item --skip-safemalloc
If MySQL is configured with @code{--with-debug=full}, all programs
will check the memory for overruns for every memory allocation and memory
freeing.  As this checking is very slow, you can avoid this, when you don't
need memory checking, by using this option.

@item --skip-show-database
Don't allow 'SHOW DATABASE' commands, unless the user has @strong{process}
privilege.

@item --skip-stack-trace
Don't write stack traces.  This option is useful when you are running
@code{mysqld} under a debugger. On some system you also have to use
this option to get a core file. @xref{Debugging server}.

@item --skip-thread-priority
Disable using thread priorities for faster response time.

@item --socket=path
Socket file to use for local connections instead of default
@code{/tmp/mysql.sock}.

@item --sql-mode=option[,option[,option...]]
Option can be any combination of: @code{REAL_AS_FLOAT},
@code{PIPES_AS_CONCAT}, @code{ANSI_QUOTES}, @code{IGNORE_SPACE},
@code{SERIALIZE}, @code{ONLY_FULL_GROUP_BY}.  It can also be empty
(@code{""}) if you want to reset this.

By specifying all of the above options is same as using --ansi.
With this option one can turn on only needed SQL modes. @xref{ANSI mode}.

@item --temp-pool
Using this option will cause most temporary files created to use a small
set of names, rather than a unique name for each new file. This is to
work around a problem in the Linux kernel dealing with creating a bunch
of new files with different names. With the old behavior, Linux seems to
'leak' memory, as it's being allocated to the directory entry cache
instead of the disk cache.

@item --transaction-isolation= @{ READ-UNCOMMITTED | READ-COMMITTED | REPEATABLE-READ | SERIALIZABLE @}
Sets the default transaction isolation level.  @xref{SET TRANSACTION}.

@item -t, --tmpdir=path
Path for temporary files. It may be useful if your default @code{/tmp}
directory resides on a partition too small to hold temporary tables.

@item -u, --user= [user_name | userid]
Run @code{mysqld} daemon as user @code{user_name} or @code{userid} (numeric).
This option is @strong{mandatory} when starting @code{mysqld} as root.

@item -V, --version
Output version information and exit.

@item -W, --warnings
Print out warnings like @code{Aborted connection...} to the @file{.err} file.
@xref{Communication errors}.
@end table


@node Option files, Installing many servers, Command-line options, Configuring MySQL
@subsection @file{my.cnf} Option Files

@cindex default options
@cindex option files
@cindex creating, default startup options
@cindex startup options, default

MySQL can, since Version 3.22, read default startup options for the
server and for clients from option files.

MySQL reads default options from the following files on Unix:

@tindex .my.cnf file
@multitable @columnfractions .35 .50
@item @strong{Filename} @tab @strong{Purpose}
@item @code{/etc/my.cnf} @tab Global options
@item @code{DATADIR/my.cnf} @tab Server-specific options
@item @code{defaults-extra-file} @tab The file specified with --defaults-extra-file=#
@item @code{~/.my.cnf} @tab User-specific options
@end multitable

@code{DATADIR} is the MySQL data directory (typically
@file{/usr/local/mysql/data} for a binary installation or
@file{/usr/local/var} for a source installation).  Note that this is the
directory that was specified at configuration time, not the one specified
with @code{--datadir} when @code{mysqld} starts up!  (@code{--datadir} has no
effect on where the server looks for option files, because it looks for them
before it processes any command-line arguments.)

MySQL reads default options from the following files on Windows:

@multitable @columnfractions .35 .50
@item @strong{Filename} @tab @strong{Purpose}
@item @code{windows-system-directory\my.ini} @tab Global options
@item @code{C:\my.cnf} @tab Global options
@end multitable

Note that on Windows, you should specify all paths with @code{/} instead of
@code{\}. If you use @code{\}, you need to specify this twice, as
@code{\} is the escape character in MySQL.

@cindex environment variables
MySQL tries to read option files in the order listed above.  If
multiple option files exist, an option specified in a file read later takes
precedence over the same option specified in a file read earlier.  Options
specified on the command-line take precedence over options specified in any
option file.  Some options can be specified using environment variables.
Options specified on the command-line or in option files take precedence over
environment variable values. @xref{Environment variables}.

The following programs support option files:  @code{mysql},
@code{mysqladmin}, @code{mysqld}, @code{mysqld_safe}, @code{mysql.server},
@code{mysqldump}, @code{mysqlimport}, @code{mysqlshow}, @code{mysqlcheck},
@code{myisamchk}, and @code{myisampack}.

Any long option that may be given on the command-line when running a MySQL
program can be given in an option file as well (without the leading double
dash).  Run the program with @code{--help} to get a list of available options.

An option file can contain lines of the following forms:

@table @code
@item #comment
Comment lines start with @samp{#} or @samp{;}. Empty lines are ignored.

@item [group]
@code{group} is the name of the program or group for which you want to set
options.  After a group line, any @code{option} or @code{set-variable} lines
apply to the named group until the end of the option file or another group
line is given.

@item option
This is equivalent to @code{--option} on the command-line.

@item option=value
This is equivalent to @code{--option=value} on the command-line.

@item set-variable = variable=value
This is equivalent to @code{--set-variable variable=value} on the command-line.
This syntax must be used to set a @code{mysqld} variable.
@end table

The @code{client} group allows you to specify options that apply to all
MySQL clients (not @code{mysqld}). This is the perfect group to use
to specify the password you use to connect to the server.  (But make
sure the option file is readable and writable only by yourself.)

Note that for options and values, all leading and trailing blanks are
automatically deleted.  You may use the escape sequences @samp{\b},
@samp{\t}, @samp{\n}, @samp{\r}, @samp{\\}, and @samp{\s} in your value string
(@samp{\s} == blank).

Here is a typical global option file:

@example
[client]
port=3306
socket=/tmp/mysql.sock

[mysqld]
port=3306
socket=/tmp/mysql.sock
set-variable = key_buffer_size=16M
set-variable = max_allowed_packet=1M

[mysqldump]
quick
@end example

Here is typical user option file:

@example
[client]
# The following password will be sent to all standard MySQL clients
password=my_password

[mysql]
no-auto-rehash
set-variable = connect_timeout=2

[mysqlhotcopy]
interactive-timeout

@end example

@tindex .my.cnf file
If you have a source distribution, you will find sample configuration
files named @file{my-xxxx.cnf} in the @file{support-files} directory.
If you have a binary distribution, look in the @file{DIR/support-files}
directory, where @code{DIR} is the pathname to the MySQL
installation directory (typically @file{/usr/local/mysql}).  Currently
there are sample configuration files for small, medium, large, and very
large systems.  You can copy @file{my-xxxx.cnf} to your home directory
(rename the copy to @file{.my.cnf}) to experiment with this.

All MySQL clients that support option files support the
following options:

@multitable @columnfractions .45 .55
@item @strong{Option} @tab @strong{Description}
@item --no-defaults @tab Don't read any option files.
@item --print-defaults @tab Print the program name and all options that it will get.
@item --defaults-file=full-path-to-default-file @tab Only use the given configuration file.
@item --defaults-extra-file=full-path-to-default-file @tab Read this configuration file after the global configuration file but before the user configuration file.
@end multitable

Note that the above options must be first on the command-line to work!
@code{--print-defaults} may however be used directly after the
@code{--defaults-xxx-file} commands.

Note for developers:  Option file handling is implemented simply by
processing all matching options (that is, options in the appropriate group)
before any command-line arguments. This works nicely for programs that use
the last instance of an option that is specified multiple times. If you have
an old program that handles multiply-specified options this way but doesn't
read option files, you need add only two lines to give it that capability.
Check the source code of any of the standard MySQL clients to see
how to do this.

In shell scripts you can use the @file{my_print_defaults} command to parse the
config files:

@example

shell> my_print_defaults client mysql
--port=3306
--socket=/tmp/mysql.sock
--no-auto-rehash
@end example

The above output contains all options for the groups 'client' and 'mysql'.


@node Installing many servers, Multiple servers, Option files, Configuring MySQL
@subsection Installing Many Servers on the Same Machine

@cindex post-install, many servers
@cindex Installing many servers
@cindex Starting many servers

In some cases you may want to have many different @code{mysqld} daemons
(servers) running on the same machine.  You may for example want to run
a new version of MySQL for testing together with an old version
that is in production.  Another case is when you want to give different
users access to different @code{mysqld} servers that they manage themselves.

One way to get a new server running is by starting it with a different
socket and port as follows:

@tindex @code{MYSQL_UNIX_PORT} environment variable
@tindex @code{MYSQL_TCP_PORT} environment variable
@tindex environment variable, @code{MYSQL_UNIX_PORT}
@tindex environment variable, @code{MYSQL_TCP_PORT}
@example
shell> MYSQL_UNIX_PORT=/tmp/mysqld-new.sock
shell> MYSQL_TCP_PORT=3307
shell> export MYSQL_UNIX_PORT MYSQL_TCP_PORT
shell> scripts/mysql_install_db
shell> bin/safe_mysqld &
@end example

The environment variables appendix includes a list of other environment
variables you can use to affect @code{mysqld}. @xref{Environment variables}.

The above is the quick and dirty way that one commonly uses for testing.
The nice thing with this is that all connections you do in the above shell
will automatically be directed to the new running server!

If you need to do this more permanently, you should create an option
file for each server. @xref{Option files}.  In your startup script that
is executed at boot time you should specify for both servers:

@code{safe_mysqld --default-file=path-to-option-file}

At least the following options should be different per server:

@itemize
@item port=#
@item socket=path
@item pid-file=path
@end itemize

The following options should be different, if they are used:

@itemize
@item log=path
@item log-bin=path
@item log-update=path
@item log-isam=path
@item bdb-logdir=path
@end itemize

If you want more performance, you can also specify the following differently:

@itemize
@item tmpdir=path
@item bdb-tmpdir=path
@end itemize

@xref{Command-line options}.

If you are installing binary MySQL versions (.tar files) and
start them with @code{./bin/safe_mysqld} then in most cases the only
option you need to add/change is the @code{socket} and @code{port}
argument to @code{safe_mysqld}.

@xref{Multiple servers, ,Running Multiple MySQL Servers on the Same Machine}.

@node Multiple servers,  , Installing many servers, Configuring MySQL
@subsection Running Multiple MySQL Servers on the Same Machine

@cindex multiple servers
@cindex servers, multiple
@cindex running, multiple servers

There are circumstances when you might want to run multiple servers on the same
machine.  For example, you might want to test a new MySQL release
while leaving your existing production setup undisturbed.  Or you might
be an Internet service provider that wants to provide independent
MySQL installations for different customers.

If you want to run multiple servers, the easiest way is to compile the servers
with different TCP/IP ports and socket files so they are not
both listening to the same TCP/IP port or socket file. @xref{mysqld_multi, ,
@code{mysqld_multi}}.

Assume an existing server is configured for the default port number and
socket file.  Then configure the new server with a @code{configure} command
something like this:

@example
shell> ./configure  --with-tcp-port=port_number \
             --with-unix-socket-path=file_name \
             --prefix=/usr/local/mysql-3.22.9
@end example

Here @code{port_number} and @code{file_name} should be different from the
default port number and socket file pathname, and the @code{--prefix} value
should specify an installation directory different from the one under which
the existing MySQL installation is located.

You can check the socket used by any currently executing MySQL server
with this command:

@example
shell> mysqladmin -h hostname --port=port_number variables
@end example

Note that if you specify ``@code{localhost}'' as a hostname, @code{mysqladmin}
will default to using Unix sockets instead of TCP/IP.

If you have a MySQL server running on the port you used, you will
get a list of some of the most important configurable variables in
MySQL, including the socket name.

You don't have to recompile a new MySQL server just to start with
a different port and socket.  You can change the port and socket to be used
by specifying them at runtime as options to @code{safe_mysqld}:

@example
shell> /path/to/safe_mysqld --socket=file_name --port=port_number
@end example

@code{mysqld_multi} can also take @code{safe_mysqld} (or @code{mysqld})
as an argument and pass the options from a configuration file to
@code{safe_mysqld} and further to @code{mysqld}.

If you run the new server on the same database directory as another
server with logging enabled, you should also specify the name of the log
files to @code{safe_mysqld} with @code{--log}, @code{--log-update}, or
@code{--log-slow-queries}.  Otherwise, both servers may be trying to
write to the same log file.

@strong{Warning}: normally you should never have two servers that update
data in the same database!  If your OS doesn't support fault-free system
locking, this may lead to unpleasant surprises!

If you want to use another database directory for the second server, you
can use the @code{--datadir=path} option to @code{safe_mysqld}.

@strong{Note} also that starting several MySQL servers
(@code{mysqlds}) in different machines and letting them access one data
directory over @code{NFS} is generally a @strong{bad idea}! The problem
is that the @code{NFS} will become the bottleneck with the speed. It is
not meant for such use. And last but not least, you would still have to
come up with a solution how to make sure that two or more @code{mysqlds}
are not interfering with each other. At the moment there is no platform
that would 100% reliable do the file locking (@code{lockd} daemon
usually) in every situation. Yet there would be one more possible risk
with @code{NFS}; it would make the work even more complicated for
@code{lockd} daemon to handle. So make it easy for your self and forget
about the idea. The working solution is to have one computer with an
operating system that efficiently handles threads and have several CPUs
in it.

When you want to connect to a MySQL server that is running with
a different port than the port that is compiled into your client, you
can use one of the following methods:

@itemize @bullet
@item
Start the client with @code{--host 'hostname' --port=port_number} to connect
with TCP/IP, or @code{[--host localhost] --socket=file_name} to connect via
a Unix socket.

@item
In your C or Perl programs, you can give the port or socket arguments
when connecting to the MySQL server.

@item
If your are using the Perl @code{DBD::mysql} module you can read the options
from the MySQL option files.  @xref{Option files}.

@example
$dsn = "DBI:mysql:test;mysql_read_default_group=client;
        mysql_read_default_file=/usr/local/mysql/data/my.cnf"
$dbh = DBI->connect($dsn, $user, $password);
@end example

@item
@tindex MYSQL_UNIX_PORT environment variable
@tindex MYSQL_TCP_PORT environment variable
@tindex environment variable, MYSQL_UNIX_PORT
@tindex environment variable, MYSQL_TCP_PORT
Set the @code{MYSQL_UNIX_PORT} and @code{MYSQL_TCP_PORT} environment variables
to point to the Unix socket and TCP/IP port before you start your clients.
If you normally use a specific socket or port, you should place commands
to set these environment variables in your @file{.login} file.
@xref{Environment variables}.

@item
@tindex .my.cnf file
Specify the default socket and TCP/IP port in the @file{.my.cnf} file in your
home directory. @xref{Option files}.
@end itemize


@node Privilege system, User Account Management, Configuring MySQL, MySQL Database Administration
@section General Security Issues and the MySQL Access Privilege System

@cindex system, security
@cindex access privileges
@cindex privileges, access
@cindex security system
@cindex ACLs

MySQL has an advanced but non-standard security/privilege
system.  This section describes how it works.

@menu
* General security::            General security
* Security::                    How to make MySQL secure against crackers
* Privileges options::          Startup Options for @code{mysqld} Concerning Security
* LOAD DATA LOCAL::             
* What Privileges::             What the privilege system does
* Privileges::                  How the privilege system works
* Privileges provided::         Privileges provided by MySQL
* Connecting::                  Connecting to the MySQL server
* Connection access::           Access control, stage 1: Connection verification
* Request access::              Access control, stage 2: Request verification
* Access denied::               Causes of @code{Access denied} errors
@end menu


@node General security, Security, Privilege system, Privilege system
@subsection General Security Guidelines

Anyone using MySQL on a computer connected to the Internet
should read this section to avoid the most common security mistakes.

In discussing security, we emphasize the necessity of fully protecting the
entire server host (not simply the MySQL server) against all types
of applicable attacks: eavesdropping, altering, playback, and denial of
service. We do not cover all aspects of availability and fault tolerance
here.

MySQL uses security based on Access Control Lists (ACLs) for all
connections, queries, and other operations that a user may attempt to
perform. There is also some support for SSL-encrypted connections
between MySQL clients and servers. Many of the concepts
discussed here are not specific to MySQL at all; the same
general ideas apply to almost all applications.

When running MySQL, follow these guidelines whenever possible:

@itemize @bullet
@item
@strong{Do not ever give anyone (except the mysql root user) access to the
@code{user} table in the @code{mysql} database!}  This is critical.
@strong{The encrypted password is the real password in MySQL.}
Anyone who knows the password which is listed in the @code{user} table
and has access to the host listed for the account @strong{can easily log
in as that user}.

@item
Learn the MySQL access privilege system. The @code{GRANT} and
@code{REVOKE} commands are used for controlling access to MySQL. Do
not grant any more privileges than necessary. Never grant privileges to all
hosts.

Checklist:
@itemize @minus
@item
Try @code{mysql -u root}. If you are able to connect successfully to the
server without being asked for a password, you have problems. Anyone
can connect to your MySQL server as the MySQL
@code{root} user with full privileges!
Review the MySQL installation instructions, paying particular
attention to the item about setting a @code{root} password.
@item
Use the command @code{SHOW GRANTS} and check to see who has access to
what. Remove those privileges that are not necessary using the @code{REVOKE}
command.
@end itemize

@item
Do not keep any plain-text passwords in your database. When your
computer becomes compromised, the intruder can take the full list of
passwords and use them. Instead use @code{MD5()} or another one-way
hashing function.
@item
Do not choose passwords from dictionaries. There are special programs to
break them. Even passwords like ``xfish98'' are very bad.  Much better is
``duag98'' which contains the same word ``fish'' but typed one key to the
left on a standard QWERTY keyboard. Another method is to use ``Mhall'' which
is taken from the first characters of each word in the sentence ``Mary had
a little lamb.'' This is easy to remember and type, but difficult to guess
for someone who does not know it.
@item
Invest in a firewall. This protects you from at least 50% of all types of
exploits in any software. Put MySQL behind the firewall or in
a demilitarised zone (DMZ).

Checklist:
@itemize @minus
@item
Try to scan your ports from the Internet using a tool such as
@code{nmap}. MySQL uses port 3306 by default. This port should
be inaccessible from untrusted hosts. Another simple way to check whether
or not your MySQL port is open is to try the following command
from some remote machine, where @code{server_host} is the hostname of
your MySQL server:

@example
shell> telnet server_host 3306
@end example

If you get a connection and some garbage characters, the port is
open, and should be closed on your firewall or router, unless you really
have a good reason to keep it open. If @code{telnet} just hangs or the
connection is refused, everything is OK; the port is blocked.
@end itemize

@item
Do not trust any data entered by your users. They can try to trick your
code by entering special or escaped character sequences in web forms,
URLs, or whatever application you have built. Be sure that your
application remains secure if a user enters something like ``@code{; DROP
DATABASE mysql;}''. This is an extreme example, but large security leaks
and data loss may occur as a result of hackers using similar techniques,
if you do not prepare for them.

Also remember to check numeric data. A common mistake is to protect only
strings. Sometimes people think that if a database contains only publicly
available data that it need not be protected. This is incorrect. At least
denial-of-service type attacks can be performed on such
databases. The simplest way to protect from this type of attack is to use
apostrophes around the numeric constants: @code{SELECT * FROM table
WHERE ID='234'} rather than @code{SELECT * FROM table WHERE ID=234}.
MySQL automatically converts this string to a number and
strips all non-numeric symbols from it.

Checklist:
@itemize @minus
@item
All web applications:
@itemize @bullet
@item
Try to enter @samp{'} and @samp{"} in all your web forms. If you get any kind
of MySQL error, investigate the problem right away.
@item
Try to modify any dynamic URLs by adding @code{%22} (@samp{"}), @code{%23}
(@samp{#}), and @code{%27} (@samp{'}) in the URL.
@item
Try to modify datatypes in dynamic URLs from numeric ones to character
ones containing characters from previous examples. Your application
should be safe against this and similar attacks.
@item
Try to enter characters, spaces, and special symbols instead of numbers in
numeric fields. Your application should remove them before passing them to
MySQL or your application should generate an error. Passing
unchecked values to MySQL is very dangerous!
@item
Check data sizes before passing them to MySQL.
@item
Consider having your application connect to the database using a
different user name than the one you use for administrative purposes. Do
not give your applications any more access privileges than they need.
@end itemize
@item
Users of PHP:
@itemize @bullet
@item Check out the @code{addslashes()} function.
As of PHP 4.0.3, a @code{mysql_escape_string()} function is available
that is based on the function of the same name in the MySQL C API.
@end itemize
@item
Users of MySQL C API:
@itemize @bullet
@item Check out the @code{mysql_real_escape_string()} API call.
@end itemize
@item
Users of MySQL++:
@itemize @bullet
@item Check out the @code{escape} and @code{quote} modifiers for query streams.
@end itemize
@item
Users of Perl DBI:
@itemize @bullet
@item Check out the @code{quote()} method or use placeholders.
@end itemize
@item
Users of Java JDBC:
@itemize @bullet
@item Use a @code{PreparedStatement} object and placeholders.
@end itemize
@end itemize

@item
Do not transmit plain (unencrypted) data over the Internet. These data are
accessible to everyone who has the time and ability to intercept it and use
it for their own purposes. Instead, use an encrypted protocol such as SSL or
SSH. MySQL supports internal SSL connections as of Version 4.0.0.
SSH port-forwarding can be used to create an encrypted (and compressed)
tunnel for the communication.
@item
Learn to use the @code{tcpdump} and @code{strings} utilities. For most cases,
you can check whether MySQL data streams are unencrypted
by issuing a command like the following:

@example
shell> tcpdump -l -i eth0 -w - src or dst port 3306 | strings
@end example

(This works under Linux and should work with small modifications under
other systems.)  Warning: If you do not see data this doesn't always
actually mean that it is encrypted. If you need high security, you should
consult with a security expert.
@end itemize


@node Security, Privileges options, General security, Privilege system
@subsection How to Make MySQL Secure Against Crackers

@cindex crackers, security against
@cindex security, against crackers

When you connect to a MySQL server, you normally should use a
password.  The password is not transmitted in clear text over the
connection, however the encryption algorithm is not very strong, and
with some effort a clever attacker can crack the password if he is able
to sniff the traffic between the client and the server. If the
connection between the client and the server goes through an untrusted
network, you should use an SSH tunnel to encrypt the
communication.

All other information is transferred as text that can be read by anyone
who is able to watch the connection.  If you are concerned about this,
you can use the compressed protocol (in MySQL Version 3.22 and above)
to make things much harder.  To make things even more secure you should use
@code{ssh}.  You can find an @code{Open Source} @code{ssh} client at
@uref{http://www.openssh.org/}, and a commercial @code{ssh} client at
@uref{http://www.ssh.com/}.  With this, you can get an encrypted TCP/IP
connection between a MySQL server and a MySQL client.

If you are using MySQL 4.0, you can also use internal openssl support.
@xref{Secure connections}.

To make a MySQL system secure, you should strongly consider the
following suggestions:

@itemize @bullet
@item
Use passwords for all MySQL users. Remember that anyone can log in
as any other person as simply as @code{mysql -u other_user db_name} if
@code{other_user} has no password.  It is common behavior with client/server
applications that the client may specify any user name.  You can change the
password of all users by editing the @code{mysql_install_db} script before
you run it, or only the password for the MySQL @code{root} user like
this:

@example
shell> mysql -u root mysql
mysql> UPDATE user SET Password=PASSWORD('new_password')
    ->             WHERE user='root';
mysql> FLUSH PRIVILEGES;
@end example

@item
Don't run the MySQL daemon as the Unix @code{root} user.  This is
very dangerous, because any user with the @strong{file} privilege will be able
to create files as @code{root} (for example, @code{~root/.bashrc}). To
prevent this, @code{mysqld} will refuse to run as @code{root} unless it
is specified directly using a @code{--user=root} option.

@code{mysqld} can be run as an ordinary unprivileged user instead.
You can also create a new Unix user @code{mysql} to make everything
even more secure.  If you run @code{mysqld} as another Unix user,
you don't need to change the @code{root} user name in the @code{user}
table, because MySQL user names have nothing to do with Unix
user names.  To start @code{mysqld} as another Unix user, add a @code{user}
line that specifies the user name to the @code{[mysqld]} group of the
@file{/etc/my.cnf} option file or the @file{my.cnf} option file in the
server's data directory. For example:

@example
[mysqld]
user=mysql
@end example

This will cause the server to start as the designated user whether you
start it manually or by using @code{safe_mysqld} or @code{mysql.server}.
For more details, see @ref{Changing MySQL user}.

@item
Don't support symlinks to tables (this can be disabled with the
@code{--skip-symlink} option). This is especially important if you run
@code{mysqld} as root as anyone that has write access to the mysqld data
directories could then delete any file in the system!
@xref{Symbolic links to tables}.

@item
Check that the Unix user that @code{mysqld} runs as is the only user with
read/write privileges in the database directories.

@item
Don't give the @strong{process} privilege to all users.  The output of
@code{mysqladmin processlist} shows the text of the currently executing
queries, so any user who is allowed to execute that command might be able to
see if another user issues an @code{UPDATE user SET
password=PASSWORD('not_secure')} query.

@code{mysqld} reserves an extra connection for users who have the
@strong{process} privilege, so that a MySQL @code{root} user can log
in and check things even if all normal connections are in use.

@item
Don't give the @strong{file} privilege to all users.  Any user that has this
privilege can write a file anywhere in the filesystem with the privileges of
the @code{mysqld} daemon!  To make this a bit safer, all files generated with
@code{SELECT ... INTO OUTFILE} are readable to everyone, and you cannot
overwrite existing files.

@tindex @file{/etc/passwd}
The @strong{file} privilege may also be used to read any file accessible
to the Unix user that the server runs as.  This could be abused, for example,
by using @code{LOAD DATA} to load @file{/etc/passwd} into a table, which
can then be read with @code{SELECT}.

@item
If you don't trust your DNS, you should use IP numbers instead of
hostnames in the grant tables. In any case, you should be very careful
about creating grant table entries using hostname values that contain
wildcards!

@item
If you want to restrict the number of connections for a single user, you
can do this by setting the @code{max_user_connections} variable in
@code{mysqld}.
@end itemize


@node Privileges options, LOAD DATA LOCAL, Security, Privilege system
@subsection Startup Options for @code{mysqld} Concerning Security

The following @code{mysqld} options affect security:

@table @code
@item --local-infile[=(0|1)]
If one uses @code{--local-infile=0} then one can't use @code{LOAD DATA LOCAL
INFILE}.

@item --safe-show-database
With this option,
@code{SHOW DATABASES} returns only those databases for which the user has
some kind of privilege.

@item --safe-user-create
If this is enabled, an user can't create new users with the @code{GRANT}
command, if the user doesn't have the @strong{insert} privilege for the
@code{mysql.user} table.  If you want to give a user access to just create
new users with those privileges that the user has right to grant, you should
give the user the following privilege:

@example
mysql> GRANT INSERT(user) ON mysql.user TO 'user'@@'hostname';
@end example

This will ensure that the user can't change any privilege columns directly,
but has to use the @code{GRANT} command to give privileges to other users.

@item --skip-grant-tables
This option causes the server not to use the privilege system at all. This
gives everyone @strong{full access} to all databases!  (You can tell a running
server to start using the grant tables again by executing @code{mysqladmin
flush-privileges} or @code{mysqladmin reload}.)

@item --skip-name-resolve
Hostnames are not resolved.  All @code{Host} column values in the grant
tables must be IP numbers or @code{localhost}.

@item --skip-networking
Don't allow TCP/IP connections over the network.  All connections to
@code{mysqld} must be made via Unix sockets. This option is unsuitable for
systems that use MIT-pthreads, because the MIT-pthreads package doesn't
support Unix sockets.

@item --skip-show-database
With this option, the
@code{SHOW DATABASES} statement doesn't return anything.

@end table

@node LOAD DATA LOCAL, What Privileges, Privileges options, Privilege system
@subsection Security issues with LOAD DATA LOCAL

In MySQL 3.23.49 and MySQL 4.0.2, we added some new options to deal with
possible security issues when it comes to @code{LOAD DATA LOCAL}.

There are two possible problems with supporting this command:

As the reading of the file is initiated from the server, one could
theoretically create a patched MySQL server that could read any file on
the client machine that the current user has read access to, when
the client issues a query against the table.

In a web environment where the clients are connecting from a web
server, a user could use @code{LOAD DATA LOCAL} to read any files
that the web server process has read access to (assuming a user could
run any command against the SQL server).

There are two separate fixes for this:

If you don't configure MySQL with @code{--enable-local-infile}, then
@code{LOAD DATA LOCAL} will be disabled by all clients, unless one
calls @code{mysql_options(... MYSQL_OPT_LOCAL_INFILE, 0)} in the client.
@xref{mysql_options, , @code{mysql_options()}}.

For the @code{mysql} command-line client, @code{LOAD DATA LOCAL} can be
enabled by specifying the option @code{--local-infile[=1]}, or disabled
with @code{--local-infile=0}.

By default, all MySQL clients and libraries are compiled with
@code{--enable-local-infile}, to be compatible with MySQL 3.23.48 and
before.

One can disable all @code{LOAD DATA LOCAL} commands in the MySQL server
by starting @code{mysqld} with @code{--local-infile=0}.

In the case that @code{LOAD DATA LOCAL INFILE} is disabled in the server or
the client, you will get the error message (1148):

@example
The used command is not allowed with this MySQL version
@end example

@node What Privileges, Privileges, LOAD DATA LOCAL, Privilege system
@subsection What the Privilege System Does

@cindex system, privilege
@cindex privilege system
@cindex passwords, security

The primary function of the MySQL privilege system is to
authenticate a user connecting from a given host, and to associate that user
with privileges on a database such as
@strong{select}, @strong{insert}, @strong{update} and @strong{delete}.

Additional functionality includes the ability to have an anonymous user and
to grant privileges for MySQL-specific functions such as @code{LOAD
DATA INFILE} and administrative operations.


@node Privileges, Privileges provided, What Privileges, Privilege system
@subsection How the Privilege System Works

@cindex privilege system, described

The MySQL privilege system ensures that all users may do exactly the
things that they are supposed to be allowed to do.  When you connect to a
MySQL server, your identity is determined by @strong{the host from
which you connect} and @strong{the user name you specify}.  The system grants
privileges according to your identity and @strong{what you want to do}.

MySQL considers both your hostname and user name in identifying you
because there is little reason to assume that a given user name belongs to
the same person everywhere on the Internet.  For example, the user
@code{joe} who connects from @code{office.com} need not be the same
person as the user @code{joe} who connects from @code{elsewhere.com}.
MySQL handles this by allowing you to distinguish users on different
hosts that happen to have the same name:  you can grant @code{joe} one set
of privileges for connections from @code{office.com}, and a different set
of privileges for connections from @code{elsewhere.com}.

MySQL access control involves two stages:

@itemize @bullet
@item
Stage 1: The server checks whether you are even allowed to connect.

@item
Stage 2: Assuming you can connect, the server checks each request you issue
to see whether you have sufficient privileges to perform it.  For
example, if you try to select rows from a table in a database or drop a table
from the database, the server makes sure you have the @strong{select}
privilege for the table or the @strong{drop} privilege for the database.
@end itemize

The server uses the @code{user}, @code{db}, and @code{host} tables in the
@code{mysql} database at both stages of access control.  The fields in these
grant tables are shown here:

@multitable @columnfractions .20 .20 .20 .20
@item @strong{Table name}       @tab @code{user}        @tab @code{db}          @tab @code{host}

@item @strong{Scope fields}     @tab @code{Host}        @tab @code{Host}        @tab @code{Host}
@item                           @tab @code{User}        @tab @code{Db}          @tab @code{Db}
@item                           @tab @code{Password}    @tab @code{User}        @tab

@item @strong{Privilege fields} @tab @code{Select_priv} @tab @code{Select_priv} @tab @code{Select_priv}
@item                           @tab @code{Insert_priv} @tab @code{Insert_priv} @tab @code{Insert_priv}
@item                           @tab @code{Update_priv} @tab @code{Update_priv} @tab @code{Update_priv}
@item                           @tab @code{Delete_priv} @tab @code{Delete_priv} @tab @code{Delete_priv}
@item                           @tab @code{Index_priv}  @tab @code{Index_priv}  @tab @code{Index_priv}
@item                           @tab @code{Alter_priv}  @tab @code{Alter_priv}  @tab @code{Alter_priv}
@item                           @tab @code{Create_priv} @tab @code{Create_priv} @tab @code{Create_priv}
@item                           @tab @code{Drop_priv}   @tab @code{Drop_priv}   @tab @code{Drop_priv}
@item                           @tab @code{Grant_priv}  @tab @code{Grant_priv}  @tab @code{Grant_priv}
@item                           @tab @code{References_priv} @tab                @tab
@item                           @tab @code{Reload_priv} @tab                    @tab
@item                           @tab @code{Shutdown_priv} @tab                  @tab
@item                           @tab @code{Process_priv} @tab                   @tab
@item                           @tab @code{File_priv}    @tab                   @tab
@end multitable

For the second stage of access control (request verification), the server
may, if the request involves tables, additionally consult the
@code{tables_priv} and @code{columns_priv} tables.  The fields in these
tables are shown here:

@multitable @columnfractions .20 .20 .20
@item @strong{Table name}   @tab @code{tables_priv} @tab @code{columns_priv}

@item @strong{Scope fields} @tab @code{Host}        @tab @code{Host}
@item                       @tab @code{Db}          @tab @code{Db}
@item                       @tab @code{User}        @tab @code{User}
@item                       @tab @code{Table_name}  @tab @code{Table_name}
@item                       @tab                    @tab @code{Column_name}

@item @strong{Privilege fields} @tab @code{Table_priv}   @tab @code{Column_priv}
@item                           @tab @code{Column_priv}  @tab

@item @strong{Other fields} @tab @code{Timestamp}   @tab @code{Timestamp}
@item                       @tab @code{Grantor}     @tab
@end multitable

Each grant table contains scope fields and privilege fields.

Scope fields determine the scope of each entry in the tables, that is, the
context in which the entry applies.  For example, a @code{user} table entry
with @code{Host} and @code{User} values of @code{'thomas.loc.gov'} and
@code{'bob'} would be used for authenticating connections made to the server
by @code{bob} from the host @code{thomas.loc.gov}.  Similarly, a @code{db}
table entry with @code{Host}, @code{User}, and @code{Db} fields of
@code{'thomas.loc.gov'}, @code{'bob'} and @code{'reports'} would be used when
@code{bob} connects from the host @code{thomas.loc.gov} to access the
@code{reports} database.  The @code{tables_priv} and @code{columns_priv}
tables contain scope fields indicating tables or table/column combinations
to which each entry applies.

@cindex case sensitivity, in access checking
For access-checking purposes, comparisons of @code{Host} values are
case-insensitive.  @code{User}, @code{Password}, @code{Db}, and
@code{Table_name} values are case-sensitive.
@code{Column_name} values are case-insensitive in MySQL Version
3.22.12 or later.

Privilege fields indicate the privileges granted by a table entry, that is,
what operations can be performed.  The server combines the information in the
various grant tables to form a complete description of a user's privileges.
The rules used to do this are described in @ref{Request access}.

Scope fields are strings, declared as shown here; the default value for
each is the empty string:

@multitable @columnfractions .20 .15 .65
@item @strong{Field name} @tab @strong{Type}     @tab @strong{Notes}
@item @code{Host}         @tab @code{CHAR(60)}   @tab
@item @code{User}         @tab @code{CHAR(16)}   @tab
@item @code{Password}     @tab @code{CHAR(16)}   @tab
@item @code{Db}           @tab @code{CHAR(64)}   @tab (@code{CHAR(60)} for the
@code{tables_priv} and @code{columns_priv} tables)
@item @code{Table_name}   @tab @code{CHAR(60)}   @tab
@item @code{Column_name}  @tab @code{CHAR(60)}   @tab
@end multitable

In the @code{user}, @code{db} and @code{host} tables,
all privilege fields are declared as @code{ENUM('N','Y')}@-each can have a
value of @code{'N'} or @code{'Y'}, and the default value is @code{'N'}.

In the @code{tables_priv} and @code{columns_priv} tables, the privilege
fields are declared as @code{SET} fields:

@multitable @columnfractions .15 .15 .55
@item @strong{Table name}
 @tab @strong{Field name}
 @tab @strong{Possible set elements}
@item @code{tables_priv}
 @tab @code{Table_priv}
 @tab @code{'Select', 'Insert', 'Update', 'Delete', 'Create', 'Drop', 'Grant', 'References', 'Index', 'Alter'}
@item @code{tables_priv}
 @tab @code{Column_priv}
 @tab @code{'Select', 'Insert', 'Update', 'References'}
@item @code{columns_priv}
 @tab @code{Column_priv}
 @tab @code{'Select', 'Insert', 'Update', 'References'}
@end multitable

Briefly, the server uses the grant tables like this:

@itemize @bullet
@item
The @code{user} table scope fields determine whether to allow or reject
incoming connections.  For allowed connections, any privileges granted in
the @code{user} table indicate the user's global (superuser) privileges.
These privileges apply to @strong{all} databases on the server.

@item
The @code{db} and @code{host} tables are used together:

@itemize @minus
@item
The @code{db} table scope fields determine which users can access which
databases from which hosts.  The privilege fields determine which operations
are allowed.

@item
The @code{host} table is used as an extension of the @code{db} table when you
want a given @code{db} table entry to apply to several hosts.  For example,
if you want a user to be able to use a database from several hosts in
your network, leave the @code{Host} value empty in the user's @code{db} table
entry, then populate the @code{host} table with an entry for each of those
hosts.  This mechanism is described more detail in @ref{Request access}.
@end itemize

@item
The @code{tables_priv} and @code{columns_priv} tables are similar to
the @code{db} table, but are more fine-grained: they apply at the
table and column levels rather than at the database level.
@end itemize

Note that administrative privileges (@strong{reload}, @strong{shutdown},
etc.) are specified only in the @code{user} table.  This is because
administrative operations are operations on the server itself and are not
database-specific, so there is no reason to list such privileges in the
other grant tables.  In fact, only the @code{user} table need
be consulted to determine whether you can perform an administrative
operation.

The @strong{file} privilege is specified only in the @code{user} table, too.
It is not an administrative privilege as such, but your ability to read or
write files on the server host is independent of the database you are
accessing.

The @code{mysqld} server reads the contents of the grant tables once, when it
starts up.  Changes to the grant tables take effect as indicated in
@ref{Privilege changes}.

When you modify the contents of the grant tables, it is a good idea to make
sure that your changes set up privileges the way you want.  For help in
diagnosing problems, see @ref{Access denied}.  For advice on security issues,
see @ref{Security}.

A useful
diagnostic tool is the @code{mysqlaccess} script, which Yves Carlier has
provided for the MySQL distribution.  Invoke @code{mysqlaccess} with
the @code{--help} option to find out how it works.
Note that @code{mysqlaccess} checks access using only the @code{user},
@code{db} and @code{host} tables.  It does not check table- or column-level
privileges.


@node Privileges provided, Connecting, Privileges, Privilege system
@subsection Privileges Provided by MySQL

@cindex privilege information, location

Information about user privileges is stored in the @code{user}, @code{db},
@code{host}, @code{tables_priv}, and @code{columns_priv} tables in the
@code{mysql} database (that is, in the database named @code{mysql}).  The
MySQL server reads the contents of these tables when it starts up
and under the circumstances indicated in @ref{Privilege changes}.

The names used in this manual to refer to the privileges provided by
MySQL are shown here, along with the table column name associated
with each privilege in the grant tables and the context in which the
privilege applies:

@multitable @columnfractions .15 .20 .35
@item @strong{Privilege}   @tab @strong{Column}      @tab @strong{Context}
@item @strong{select}      @tab @code{Select_priv}   @tab tables
@item @strong{insert}      @tab @code{Insert_priv}   @tab tables
@item @strong{update}      @tab @code{Update_priv}   @tab tables
@item @strong{delete}      @tab @code{Delete_priv}   @tab tables
@item @strong{index}       @tab @code{Index_priv}    @tab tables
@item @strong{alter}       @tab @code{Alter_priv}    @tab tables
@item @strong{create}      @tab @code{Create_priv}   @tab databases, tables, or indexes
@item @strong{drop}        @tab @code{Drop_priv}     @tab databases or tables
@item @strong{grant}       @tab @code{Grant_priv}    @tab databases or tables
@item @strong{references}  @tab @code{References_priv} @tab databases or tables
@item @strong{reload}      @tab @code{Reload_priv}   @tab server administration
@item @strong{shutdown}    @tab @code{Shutdown_priv} @tab server administration
@item @strong{process}     @tab @code{Process_priv}  @tab server administration
@item @strong{file}        @tab @code{File_priv}     @tab file access on server
@end multitable

The @strong{select}, @strong{insert}, @strong{update}, and @strong{delete}
privileges allow you to perform operations on rows in existing tables in
a database.

@code{SELECT} statements require the @strong{select} privilege only if they
actually retrieve rows from a table.  You can execute certain @code{SELECT}
statements even without permission to access any of the databases on the
server.  For example, you could use the @code{mysql} client as a simple
calculator:

@example
mysql> SELECT 1+1;
mysql> SELECT PI()*2;
@end example

The @strong{index} privilege allows you to create or drop (remove) indexes.

The @strong{alter} privilege allows you to use @code{ALTER TABLE}.

The @strong{create} and @strong{drop} privileges allow you to create new
databases and tables, or to drop (remove) existing databases and tables.

Note that if you grant the @strong{drop} privilege for the @code{mysql}
database to a user, that user can drop the database in which the
MySQL access privileges are stored!

The @strong{grant} privilege allows you to give to other users those
privileges you yourself possess.

The @strong{file} privilege gives you permission to read and write files on
the server using the @code{LOAD DATA INFILE} and @code{SELECT ... INTO
OUTFILE} statements.  Any user to whom this privilege is granted can read or
write any file that the MySQL server can read or write.

The remaining privileges are used for administrative operations, which are
performed using the @code{mysqladmin} program.  The table here shows which
@code{mysqladmin} commands each administrative privilege allows you to
execute:

@multitable @columnfractions .15 .75
@item @strong{Privilege}  @tab @strong{Commands permitted to privilege holders}
@item @strong{reload}     @tab @code{reload}, @code{refresh},
@code{flush-privileges}, @code{flush-hosts}, @code{flush-logs}, and
@code{flush-tables}
@item @strong{shutdown}   @tab @code{shutdown}
@item @strong{process}    @tab @code{processlist}, @code{kill}
@end multitable

The @code{reload} command tells the server to re-read the grant tables.  The
@code{refresh} command flushes all tables and opens and closes the log
files.  @code{flush-privileges} is a synonym for @code{reload}.  The other
@code{flush-*} commands perform functions similar to @code{refresh} but are
more limited in scope, and may be preferable in some instances.  For example,
if you want to flush just the log files, @code{flush-logs} is a better choice
than @code{refresh}.

The @code{shutdown} command shuts down the server.

The @code{processlist} command displays information about the threads
executing within the server.  The @code{kill} command kills server threads.
You can always display or kill your own threads, but you need the
@strong{process} privilege to display or kill threads initiated by other
users. @xref{KILL}.

It is a good idea in general to grant privileges only to those users who need
them, but you should exercise particular caution in granting certain
privileges:

@itemize @bullet
@item
The @strong{grant} privilege allows users to give away their privileges to
other users.  Two users with different privileges and with the @strong{grant}
privilege are able to combine privileges.

@item
The @strong{alter} privilege may be used to subvert the privilege system
by renaming tables.

@item
The @strong{file} privilege can be abused to read any world-readable file on
the server into a database table, the contents of which can then be
accessed using @code{SELECT}. This includes the contents of all databases
hosted by the server!

@item
The @strong{shutdown} privilege can be abused to deny service to other
users entirely, by terminating the server.

@item
The @strong{process} privilege can be used to view the plain text of
currently executing queries, including queries that set or change passwords.

@item
Privileges on the @code{mysql} database can be used to change passwords
and other access privilege information.  (Passwords are stored
encrypted, so a malicious user cannot simply read them to know the plain
text password.)  If they can access the @code{mysql.user} password
column, they can use it to log into the MySQL server
for the given user.  (With sufficient privileges, the same user can
replace a password with a different one.)
@end itemize

There are some things that you cannot do with the MySQL
privilege system:

@itemize @bullet
@item
You cannot explicitly specify that a given user should be denied access.
That is, you cannot explicitly match a user and then refuse the connection.

@item
You cannot specify that a user has privileges to create or drop tables
in a database but not to create or drop the database itself.
@end itemize


@node Connecting, Connection access, Privileges provided, Privilege system
@subsection Connecting to the MySQL Server

@cindex connecting, to the server
@cindex default hostname
@cindex hostname, default
@cindex server, connecting

MySQL client programs generally require that you specify connection
parameters when you want to access a MySQL server: the host you want
to connect to, your user name, and your password.  For example, the
@code{mysql} client can be started like this (optional arguments are enclosed
between @samp{[} and @samp{]}):

@example
shell> mysql [-h host_name] [-u user_name] [-pyour_pass]
@end example

Alternate forms of the @code{-h}, @code{-u}, and @code{-p} options are
@code{--host=host_name}, @code{--user=user_name}, and
@code{--password=your_pass}.  Note that there is @emph{no space} between
@code{-p} or @code{--password=} and the password following it.

@strong{Note}: Specifying a password on the command-line is not secure!
Any user on your system may then find out your password by typing a command
like: @code{ps auxww}.  @xref{Option files}.

@code{mysql} uses default values for connection parameters that are missing
from the command-line:

@itemize @bullet
@item
The default hostname is @code{localhost}.

@item
The default user name is your Unix login name.

@item
No password is supplied if @code{-p} is missing.
@end itemize

Thus, for a Unix user @code{joe}, the following commands are equivalent:

@example
shell> mysql -h localhost -u joe
shell> mysql -h localhost
shell> mysql -u joe
shell> mysql
@end example

Other MySQL clients behave similarly.

On Unix systems, you can specify different default values to be used when you
make a connection, so that you need not enter them on the command-line each
time you invoke a client program.  This can be done in a couple of ways:

@itemize @bullet
@item
@tindex .my.cnf file
You can specify connection parameters in the @code{[client]} section of the
@file{.my.cnf} configuration file in your home directory.  The relevant
section of the file might look like this:

@example
[client]
host=host_name
user=user_name
password=your_pass
@end example

@xref{Option files}.

@item
@tindex @code{MYSQL_HOST} environment variable
@tindex environment variable, @code{MYSQL_HOST}
@tindex @code{MYSQL_PWD} environment variable
@tindex environment variable, @code{MYSQL_PWD}
@tindex @code{USER} environment variable
@tindex environment variable, @code{USER}
You can specify connection parameters using environment variables.  The
host can be specified for @code{mysql} using @code{MYSQL_HOST}.  The
MySQL user name can be specified using @code{USER} (this is for
Windows only).  The password can be specified using @code{MYSQL_PWD}
(but this is insecure; see the next section).  @xref{Environment variables}.
@end itemize


@node Connection access, Request access, Connecting, Privilege system
@subsection Access Control, Stage 1: Connection Verification

@cindex access control
@cindex control access
@cindex connecting, verification
@cindex testing, connection to the server

When you attempt to connect to a MySQL server, the server accepts or
rejects the connection based on your identity and whether you can
verify your identity by supplying the correct password.  If not, the server
denies access to you completely.  Otherwise, the server accepts the
connection, then enters Stage 2 and waits for requests.

Your identity is based on two pieces of information:

@itemize @bullet
@item
The host from which you connect

@item
Your MySQL user name
@end itemize

Identity checking is performed using the three @code{user} table scope fields
(@code{Host}, @code{User}, and @code{Password}).  The server accepts the
connection only if a @code{user} table entry matches your hostname and user
name, and you supply the correct password.

Values in the @code{user} table scope fields may be specified as follows:

@itemize @bullet
@item
A @code{Host} value may be a hostname or an IP number, or @code{'localhost'}
to indicate the local host.

@item
@cindex wildcards, in @code{mysql.user} table
You can use the wildcard characters @samp{%} and @samp{_} in the @code{Host}
field.

@item
A @code{Host} value of @code{'%'} matches any hostname.

@item
A blank @code{Host} value means that the privilege should be anded
with the entry in the @code{host} table that matches the given host name.
You can find more information about this in the next chapter.

@cindex netmask notation, in @code{mysql.user} table
@item
As of MySQL Version 3.23, for @code{Host} values specified as
IP numbers, you can specify a netmask indicating how many address bits to
use for the network number. For example:

@example
mysql> GRANT ALL PRIVILEGES ON db.*
    -> TO david@@'192.58.197.0/255.255.255.0';
@end example

This will allow everyone to connect from an IP where the following is true:

@example
user_ip & netmask = host_ip.
@end example

In the above example all IP:s in the interval 192.58.197.0 -
192.58.197.255 can connect to the MySQL server.

@item
@cindex anonymous user
Wildcard characters are not allowed in the @code{User} field, but you can
specify a blank value, which matches any name.  If the @code{user} table
entry that matches an incoming connection has a blank user name, the user is
considered to be the anonymous user (the user with no name), rather than the
name that the client actually specified.  This means that a blank user name
is used for all further access checking for the duration of the connection
(that is, during Stage 2).

@item
The @code{Password} field can be blank.  This does not mean that any password
matches, it means the user must connect without specifying a password.
@end itemize

@findex PASSWORD()
Non-blank @code{Password} values represent encrypted passwords.
MySQL does not store passwords in plaintext form for anyone to
see.  Rather, the password supplied by a user who is attempting to
connect is encrypted (using the @code{PASSWORD()} function). The
encrypted password is then used when the client/server is checking if
the password is correct. (This is done without the encrypted password
ever traveling over the connection.)  Note that from MySQL's
point of view the encrypted password is the REAL password, so you should
not give anyone access to it!  In particular, don't give normal users
read access to the tables in the @code{mysql} database!

The examples here show how various combinations of @code{Host} and
@code{User} values in @code{user} table entries apply to incoming
connections:

@multitable @columnfractions .35 .15 .50
@item @code{Host} @strong{value} @tab @code{User} @strong{value} @tab @strong{Connections matched by entry}
@item @code{'thomas.loc.gov'} @tab @code{'fred'} @tab @code{fred}, connecting from @code{thomas.loc.gov}
@item @code{'thomas.loc.gov'} @tab @code{''} @tab Any user, connecting from @code{thomas.loc.gov}
@item  @code{'%'} @tab @code{'fred'} @tab @code{fred}, connecting from any host
@item @code{'%'} @tab  @code{''} @tab Any user, connecting from any host
@item @code{'%.loc.gov'} @tab @code{'fred'} @tab @code{fred}, connecting from any host in the @code{loc.gov} domain
@item @code{'x.y.%'} @tab @code{'fred'} @tab @code{fred}, connecting from @code{x.y.net}, @code{x.y.com},@code{x.y.edu}, etc. (this is probably not useful)
@item @code{'144.155.166.177'} @tab @code{'fred'} @tab @code{fred}, connecting from the host with IP address @code{144.155.166.177}
@item @code{'144.155.166.%'} @tab @code{'fred'} @tab @code{fred}, connecting from any host in the @code{144.155.166} class C subnet
@item @code{'144.155.166.0/255.255.255.0'} @tab @code{'fred'} @tab Same as previous example
@end multitable

Because you can use IP wildcard values in the @code{Host} field (for example,
@code{'144.155.166.%'} to match every host on a subnet), there is the
possibility that someone might try to exploit this capability by naming a
host @code{144.155.166.somewhere.com}.  To foil such attempts, MySQL
disallows matching on hostnames that start with digits and a dot. Thus, if
you have a host named something like @code{1.2.foo.com}, its name will never
match the @code{Host} column of the grant tables.  Only an IP number can
match an IP wildcard value.

An incoming connection may be matched by more than one entry in the
@code{user} table.  For example, a connection from @code{thomas.loc.gov} by
@code{fred} would be matched by several of the entries just shown above.  How
does the server choose which entry to use if more than one matches?  The
server resolves this question by sorting the @code{user} table after reading
it at startup time, then looking through the entries in sorted order when a
user attempts to connect.  The first matching entry is the one that is used.

@code{user} table sorting works as follows.  Suppose the @code{user} table
looks like this:

@example
+-----------+----------+-
| Host      | User     | ...
+-----------+----------+-
| %         | root     | ...
| %         | jeffrey  | ...
| localhost | root     | ...
| localhost |          | ...
+-----------+----------+-
@end example

When the server reads in the table, it orders the entries with the
most-specific @code{Host} values first (@code{'%'} in the @code{Host} column
means ``any host'' and is least specific).  Entries with the same @code{Host}
value are ordered with the most-specific @code{User} values first (a blank
@code{User} value means ``any user'' and is least specific).  The resulting
sorted @code{user} table looks like this:

@example
+-----------+----------+-
| Host      | User     | ...
+-----------+----------+-
| localhost | root     | ...
| localhost |          | ...
| %         | jeffrey  | ...
| %         | root     | ...
+-----------+----------+-
@end example

@cindex grant tables, sorting
@cindex sorting, grant tables
@cindex @code{user} table, sorting
When a connection is attempted, the server looks through the sorted entries
and uses the first match found.  For a connection from @code{localhost} by
@code{jeffrey}, the entries with @code{'localhost'} in the @code{Host} column
match first.  Of those, the entry with the blank user name matches both the
connecting hostname and user name.  (The @code{'%'/'jeffrey'} entry would
have matched, too, but it is not the first match in the table.)

Here is another example.  Suppose the @code{user} table looks like this:

@example
+----------------+----------+-
| Host           | User     | ...
+----------------+----------+-
| %              | jeffrey  | ...
| thomas.loc.gov |          | ...
+----------------+----------+-
@end example

The sorted table looks like this:

@example
+----------------+----------+-
| Host           | User     | ...
+----------------+----------+-
| thomas.loc.gov |          | ...
| %              | jeffrey  | ...
+----------------+----------+-
@end example

A connection from @code{thomas.loc.gov} by @code{jeffrey} is matched by the
first entry, whereas a connection from @code{whitehouse.gov} by
@code{jeffrey} is matched by the second.

A common misconception is to think that for a given user name, all entries
that explicitly name that user will be used first when the server attempts to
find a match for the connection.  This is simply not true.  The previous
example illustrates this, where a connection from @code{thomas.loc.gov} by
@code{jeffrey} is first matched not by the entry containing @code{'jeffrey'}
as the @code{User} field value, but by the entry with no user name!

If you have problems connecting to the server, print out the @code{user}
table and sort it by hand to see where the first match is being made.


@node Request access, Access denied, Connection access, Privilege system
@subsection Access Control, Stage 2: Request Verification

Once you establish a connection, the server enters Stage 2.  For each request
that comes in on the connection, the server checks whether you have
sufficient privileges to perform it, based on the type of operation you wish
to perform.  This is where the privilege fields in the grant tables come into
play.  These privileges can come from any of the @code{user}, @code{db},
@code{host}, @code{tables_priv}, or @code{columns_priv} tables.  The grant
tables are manipulated with @code{GRANT} and @code{REVOKE} commands.
@xref{GRANT, , @code{GRANT}}.  (You may find it helpful to refer to
@ref{Privileges}, which lists the fields present in each of the grant
tables.)

The @code{user} table grants privileges that are assigned to you on a global
basis and that apply no matter what the current database is.  For example, if
the @code{user} table grants you the @strong{delete} privilege, you can
delete rows from any database on the server host!  In other words,
@code{user} table privileges are superuser privileges.  It is wise to grant
privileges in the @code{user} table only to superusers such as server or
database administrators.  For other users, you should leave the privileges
in the @code{user} table set to @code{'N'} and grant privileges on a
database-specific basis only, using the @code{db} and @code{host} tables.

@cindex anonymous user
@cindex wildcards, in @code{mysql.db} table
@cindex wildcards, in @code{mysql.host} table
The @code{db} and @code{host} tables grant database-specific privileges.
Values in the scope fields may be specified as follows:

@itemize @bullet
@item
The wildcard characters @samp{%} and @samp{_} can be used in the @code{Host}
and @code{Db} fields of either table.

@item
A @code{'%'} @code{Host} value in the @code{db} table means ``any host.'' A
blank @code{Host} value in the @code{db} table means ``consult the
@code{host} table for further information.''

@item
A @code{'%'} or blank @code{Host} value in the @code{host} table means ``any
host.''

@item
A @code{'%'} or blank @code{Db} value in either table means ``any database.''

@item
A blank @code{User} value in either table matches the anonymous user.
@end itemize

@cindex grant tables, sorting
@cindex sorting, grant tables
@cindex @code{db} table, sorting
@cindex @code{host} table, sorting
The @code{db} and @code{host} tables are read in and sorted when the server
starts up (at the same time that it reads the @code{user} table).  The
@code{db} table is sorted on the @code{Host}, @code{Db}, and @code{User} scope
fields, and the @code{host} table is sorted on the @code{Host} and @code{Db}
scope fields.  As with the @code{user} table, sorting puts the most-specific
values first and least-specific values last, and when the server looks for
matching entries, it uses the first match that it finds.


@cindex wildcards, in @code{mysql.tables_priv} table
@cindex wildcards, in @code{mysql.columns_priv} table
The @code{tables_priv} and @code{columns_priv} tables grant table- and
column-specific privileges.  Values in the scope fields may be specified as
follows:

@itemize @bullet
@item
The wildcard characters @samp{%} and @samp{_}
can be used in the @code{Host} field of either table.

@item
A @code{'%'} or blank @code{Host} value in either table means ``any host.''

@item
The @code{Db}, @code{Table_name} and @code{Column_name} fields cannot contain
wildcards or be blank in either table.
@end itemize

The @code{tables_priv} and @code{columns_priv} tables are sorted on
the @code{Host}, @code{Db}, and @code{User} fields.  This is similar to
@code{db} table sorting, although the sorting is simpler because
only the @code{Host} field may contain wildcards.

The request verification process is described here.  (If you are familiar
with the access-checking source code, you will notice that the description
here differs slightly from the algorithm used in the code.  The description
is equivalent to what the code actually does; it differs only to make the
explanation simpler.)

For administrative requests (@strong{shutdown}, @strong{reload}, etc.), the
server checks only the @code{user} table entry, because that is the only table
that specifies administrative privileges.  Access is granted if the entry
allows the requested operation and denied otherwise.  For example, if you
want to execute @code{mysqladmin shutdown} but your @code{user} table entry
doesn't grant the @strong{shutdown} privilege to you, access is denied
without even checking the @code{db} or @code{host} tables.  (They
contain no @code{Shutdown_priv} column, so there is no need to do so.)

For database-related requests (@strong{insert}, @strong{update}, etc.), the
server first checks the user's global (superuser) privileges by looking in
the @code{user} table entry.  If the entry allows the requested operation,
access is granted.  If the global privileges in the @code{user} table are
insufficient, the server determines the user's database-specific privileges
by checking the @code{db} and @code{host} tables:

@enumerate
@item
The server looks in the @code{db} table for a match on the @code{Host},
@code{Db}, and @code{User} fields.  The @code{Host} and @code{User} fields are
matched to the connecting user's hostname and MySQL user name.  The
@code{Db} field is matched to the database the user wants to access.  If
there is no entry for the @code{Host} and @code{User}, access is denied.

@item
If there is a matching @code{db} table entry and its @code{Host} field is
not blank, that entry defines the user's database-specific privileges.

@item
If the matching @code{db} table entry's @code{Host} field is blank, it
signifies that the @code{host} table enumerates which hosts should be allowed
access to the database.  In this case, a further lookup is done in the
@code{host} table to find a match on the @code{Host} and @code{Db} fields.
If no @code{host} table entry matches, access is denied.  If there is a
match, the user's database-specific privileges are computed as the
intersection (@strong{not} the union!) of the privileges in the @code{db} and
@code{host} table entries, that is, the privileges that are @code{'Y'} in both
entries.  (This way you can grant general privileges in the @code{db} table
entry and then selectively restrict them on a host-by-host basis using the
@code{host} table entries.)
@end enumerate

After determining the database-specific privileges granted by the @code{db}
and @code{host} table entries, the server adds them to the global privileges
granted by the @code{user} table.  If the result allows the requested
operation, access is granted.  Otherwise, the server checks the user's
table and column privileges in the @code{tables_priv} and @code{columns_priv}
tables and adds those to the user's privileges.  Access is allowed or denied
based on the result.

Expressed in boolean terms, the preceding description of how a user's
privileges are calculated may be summarised like this:

@example
global privileges
OR (database privileges AND host privileges)
OR table privileges
OR column privileges
@end example

It may not be apparent why, if the global @code{user} entry privileges are
initially found to be insufficient for the requested operation, the server
adds those privileges to the database-, table-, and column-specific privileges
later. The reason is that a request might require more than one type of
privilege.  For example, if you execute an @code{INSERT ...  SELECT}
statement, you need both @strong{insert} and @strong{select} privileges.
Your privileges might be such that the @code{user} table entry grants one
privilege and the @code{db} table entry grants the other.  In this case, you
have the necessary privileges to perform the request, but the server cannot
tell that from either table by itself; the privileges granted by the entries
in both tables must be combined.

@cindex @code{host} table
@cindex tables, @code{host}

The @code{host} table can be used to maintain a list of secure servers.

At TcX, the @code{host} table contains a list of all machines on the local
network. These are granted all privileges.

You can also use the @code{host} table to indicate hosts that are @strong{not}
secure.  Suppose you have a machine @code{public.your.domain} that is located
in a public area that you do not consider secure.  You can allow access to
all hosts on your network except that machine by using @code{host} table
entries
like this:

@example
+--------------------+----+-
| Host               | Db | ...
+--------------------+----+-
| public.your.domain | %  | ... (all privileges set to 'N')
| %.your.domain      | %  | ... (all privileges set to 'Y')
+--------------------+----+-
@end example

@cindex privilege, changes
@cindex changes to privileges
@cindex tables, grant
@cindex grant tables

Naturally, you should always test your entries in the grant tables (for
example, using @code{mysqlaccess}) to make sure your access privileges are
actually set up the way you think they are.


@node Access denied,  , Request access, Privilege system
@subsection Causes of @code{Access denied} Errors

If you encounter @code{Access denied} errors when you try to connect to the
MySQL server, the following list indicates some courses of
action you can take to correct the problem:

@itemize @bullet
@item
After installing MySQL, did you run the @code{mysql_install_db}
script to set up the initial grant table contents?  If not, do so.
@xref{Default privileges}.  Test the initial privileges by executing
this command:

@example
shell> mysql -u root test
@end example

The server should let you connect without error.  You should also make sure
you have a file @file{user.MYD} in the MySQL database directory.
Ordinarily, this is @file{PATH/var/mysql/user.MYD}, where @code{PATH} is the
pathname to the MySQL installation root.

@item
After a fresh installation, you should connect to the server and set up
your users and their access permissions:

@example
shell> mysql -u root mysql
@end example

The server should let you connect because the MySQL @code{root} user
has no password initially.  That is also a security risk, so setting the
@code{root} password is something you should do while you're setting up
your other MySQL users.

If you try to connect as @code{root} and get this error:

@example
Access denied for user: '@@unknown' to database mysql
@end example

this means that you don't have an entry in the @code{user} table with a
@code{User} column value of @code{'root'} and that @code{mysqld} cannot
resolve the hostname for your client.  In this case, you must restart the
server with the @code{--skip-grant-tables} option and edit your
@file{/etc/hosts} or @file{\windows\hosts} file to add an entry for your
host.

@item
If you get an error like the following:

@example
shell> mysqladmin -u root -pxxxx ver
Access denied for user: 'root@@localhost' (Using password: YES)
@end example

It means that you are using a wrong password. @xref{Passwords}.

If you have forgot the root password, you can restart @code{mysqld} with
@code{--skip-grant-tables} to change the password. You can find more
about this option later on in this manual section.

If you get the above error even if you haven't specified a password,
this means that you a wrong password in some @code{my.ini}
file. @xref{Option files}.  You can avoid using option files with the @code{--no-defaults} option, as follows:

@example
shell> mysqladmin --no-defaults -u root ver
@end example

@item
@cindex @code{mysql_fix_privilege_tables}
If you updated an existing MySQL installation from a version earlier
than Version 3.22.11 to Version 3.22.11 or later, did you run the
@code{mysql_fix_privilege_tables} script?  If not, do so.  The structure of
the grant tables changed with MySQL Version 3.22.11 when the
@code{GRANT} statement became functional.

@item
If your privileges seem to have changed in the middle of a session, it may be
that a superuser has changed them.  Reloading the grant tables affects new
client connections, but it also affects existing connections as indicated in
@ref{Privilege changes}.

@item
If you can't get your password to work, remember that you must use
the @code{PASSWORD()} function if you set the password with the
@code{INSERT}, @code{UPDATE}, or @code{SET PASSWORD} statements.  The
@code{PASSWORD()} function is unnecessary if you specify the password using
the @code{GRANT ... INDENTIFIED BY} statement or the @code{mysqladmin
password} command.
@xref{Passwords}.

@item
@code{localhost} is a synonym for your local hostname, and is also the
default host to which clients try to connect if you specify no host
explicitly.  However, connections to @code{localhost} do not work if you are
running on a system that uses MIT-pthreads (@code{localhost} connections are
made using Unix sockets, which are not supported by MIT-pthreads).  To avoid
this problem on such systems, you should use the @code{--host} option to name
the server host explicitly.  This will make a TCP/IP connection to the
@code{mysqld} server.  In this case, you must have your real hostname in
@code{user} table entries on the server host.  (This is true even if you are
running a client program on the same host as the server.)

@item
If you get an @code{Access denied} error when trying to connect to the
database with @code{mysql -u user_name db_name}, you may have a problem
with the @code{user} table. Check this by executing @code{mysql -u root
mysql} and issuing this SQL statement:

@example
mysql> SELECT * FROM user;
@end example

The result should include an entry with the @code{Host} and @code{User}
columns matching your computer's hostname and your MySQL user name.

@item
The @code{Access denied} error message will tell you who you are trying
to log in as, the host from which you are trying to connect, and whether
or not you were using a password. Normally, you should have one entry in
the @code{user} table that exactly matches the hostname and user name
that were given in the error message. For example if you get an error
message that contains @code{Using password: NO}, this means that you
tried to login without an password.

@item
If you get the following error when you try to connect from a different host
than the one on which the MySQL server is running, then there is no
row in the @code{user} table that matches that host:

@example
Host ... is not allowed to connect to this MySQL server
@end example

You can fix this by using the command-line tool @code{mysql} (on the
server host!) to add a row to the @code{user}, @code{db}, or @code{host}
table for the user/hostname combination from which you are trying to
connect and then execute @code{mysqladmin flush-privileges}.  If you are
not running MySQL Version 3.22 and you don't know the IP number or
hostname of the machine from which you are connecting, you should put an
entry with @code{'%'} as the @code{Host} column value in the @code{user}
table and restart @code{mysqld} with the @code{--log} option on the
server machine.  After trying to connect from the client machine, the
information in the MySQL log will indicate how you really did
connect.  (Then replace the @code{'%'} in the @code{user} table entry
with the actual hostname that shows up in the log.  Otherwise, you'll
have a system that is insecure.)

Another reason for this error on Linux is that you are using a binary
MySQL version that is compiled with a different glibc version
than the one you are using.  In this case you should either upgrade your
OS/glibc or download the source MySQL version and compile this
yourself.  A source RPM is normally trivial to compile and install, so
this isn't a big problem.

@item
If you get an error message where the hostname is not shown or where the
hostname is an IP, even if you try to connect with a hostname:

@example
shell> mysqladmin -u root -pxxxx -h some-hostname ver
Access denied for user: 'root@@' (Using password: YES)
@end example

This means that MySQL got some error when trying to resolve the
IP to a hostname.  In this case you can execute @code{mysqladmin
flush-hosts} to reset the internal DNS cache. @xref{DNS}.

Some permanent solutions are:

@itemize @minus
@item
Try to find out what is wrong with your DNS server and fix this.

@item
Specify IPs instead of hostnames in the MySQL privilege tables.

@item
Start @code{mysqld} with @code{--skip-name-resolve}.

@item
Start @code{mysqld} with @code{--skip-host-cache}.

@item
Connect to @code{localhost} if you are running the server and the client
on the same machine.

@item
Put the client machine names in @code{/etc/hosts}.
@end itemize

@item
If @code{mysql -u root test} works but @code{mysql -h your_hostname -u root
test} results in @code{Access denied}, then you may not have the correct name
for your host in the @code{user} table.  A common problem here is that the
@code{Host} value in the user table entry specifies an unqualified hostname,
but your system's name resolution routines return a fully qualified domain
name (or vice-versa).  For example, if you have an entry with host
@code{'tcx'} in the @code{user} table, but your DNS tells MySQL that
your hostname is @code{'tcx.subnet.se'}, the entry will not work. Try adding
an entry to the @code{user} table that contains the IP number of your host as
the @code{Host} column value.  (Alternatively, you could add an entry to the
@code{user} table with a @code{Host} value that contains a wildcard---for
example, @code{'tcx.%'}.  However, use of hostnames ending with @samp{%} is
@strong{insecure} and is @strong{not} recommended!)

@item
If @code{mysql -u user_name test} works but @code{mysql -u user_name
other_db_name} doesn't work, you don't have an entry for @code{other_db_name}
listed in the @code{db} table.

@item
If @code{mysql -u user_name db_name} works when executed on the server
machine, but @code{mysql -u host_name -u user_name db_name} doesn't work when
executed on another client machine, you don't have the client machine listed
in the @code{user} table or the @code{db} table.

@item
If you can't figure out why you get @code{Access denied}, remove from the
@code{user} table all entries that have @code{Host} values containing
wildcards (entries that contain @samp{%} or @samp{_}).  A very common error
is to insert a new entry with @code{Host}=@code{'%'} and
@code{User}=@code{'some user'}, thinking that this will allow you to specify
@code{localhost} to connect from the same machine.  The reason that this
doesn't work is that the default privileges include an entry with
@code{Host}=@code{'localhost'} and @code{User}=@code{''}.  Because that entry
has a @code{Host} value @code{'localhost'} that is more specific than
@code{'%'}, it is used in preference to the new entry when connecting from
@code{localhost}!  The correct procedure is to insert a second entry with
@code{Host}=@code{'localhost'} and @code{User}=@code{'some_user'}, or to
remove the entry with @code{Host}=@code{'localhost'} and
@code{User}=@code{''}.

@item
If you get the following error, you may have a problem with the @code{db} or
@code{host} table:

@example
Access to database denied
@end example

If the entry selected from the @code{db} table has an empty value in the
@code{Host} column, make sure there are one or more corresponding entries in
the @code{host} table specifying which hosts the @code{db} table entry
applies to.

If you get the error when using the SQL commands @code{SELECT ...
INTO OUTFILE} or @code{LOAD DATA INFILE}, your entry in the @code{user} table
probably doesn't have the @strong{file} privilege enabled.

@item
@cindex configuration files
@cindex environment variables
@tindex .my.cnf file
Remember that client programs will use connection parameters specified
in configuration files or environment variables.  @xref{Environment
variables}.  If a client seems to be sending the wrong default
connection parameters when you don't specify them on the command-line,
check your environment and the @file{.my.cnf} file in your home
directory.  You might also check the system-wide MySQL
configuration files, though it is far less likely that client connection
parameters will be specified there. @xref{Option files}.  If you get
@code{Access denied} when you run a client without any options, make
sure you haven't specified an old password in any of your option files!
@xref{Option files}.

@item
If you make changes to the grant tables directly (using an @code{INSERT} or
@code{UPDATE} statement) and your changes seem to be ignored, remember
that you must issue a @code{FLUSH PRIVILEGES} statement or execute a
@code{mysqladmin flush-privileges} command to cause the server to re-read
the privilege tables. Otherwise, your changes have no effect until the
next time the server is restarted.  Remember that after you set the
@code{root} password with an @code{UPDATE} command, you won't need to
specify it until after you flush the privileges, because the server
won't know you've changed the password yet!

@item
If you have access problems with a Perl, PHP, Python, or ODBC program, try to
connect to the server with @code{mysql -u user_name db_name} or @code{mysql
-u user_name -pyour_pass db_name}.  If you are able to connect using the
@code{mysql} client, there is a problem with your program and not with the
access privileges.  (Note that there is no space between @code{-p} and the
password; you can also use the @code{--password=your_pass} syntax to specify
the password. If you use the @code{-p} option alone, MySQL will
prompt you for the password.)

@item
For testing, start the @code{mysqld} daemon with the
@code{--skip-grant-tables} option.  Then you can change the MySQL
grant tables and use the @code{mysqlaccess} script to check whether
your modifications have the desired effect.  When you are satisfied with your
changes, execute @code{mysqladmin flush-privileges} to tell the @code{mysqld}
server to start using the new grant tables.  @strong{Note}: reloading the
grant tables overrides the @code{--skip-grant-tables} option.  This allows
you to tell the server to begin using the grant tables again without bringing
it down and restarting it.

@item
If everything else fails, start the @code{mysqld} daemon with a debugging
option (for example, @code{--debug=d,general,query}). This will print host and
user information about attempted connections, as well as information about
each command issued. @xref{Making trace files}.

@item
If you have any other problems with the MySQL grant tables and
feel you must post the problem to the mailing list, always provide a
dump of the MySQL grant tables. You can dump the tables with
the @code{mysqldump mysql} command. As always, post your problem using
the @code{mysqlbug} script.  @xref{Bug reports}.  In some cases you may need
to restart @code{mysqld} with @code{--skip-grant-tables} to run
@code{mysqldump}.
@end itemize


@node User Account Management, Disaster Prevention, Privilege system, MySQL Database Administration
@section MySQL User Account Management

@c FIX: add section intro

@menu
* GRANT::                       @code{GRANT} and @code{REVOKE} Syntax
* User names::                  MySQL User Names and Passwords
* Privilege changes::           When Privilege Changes Take Effect
* Default privileges::          Setting Up the Initial MySQL Privileges
* Adding users::                Adding New Users to MySQL
* User resources::              Limiting user resources
* Passwords::                   Setting Up Passwords
* Password security::           Keeping Your Password Secure
* Secure connections::          Using Secure Connections
@end menu


@node GRANT, User names, User Account Management, User Account Management
@subsection @code{GRANT} and @code{REVOKE} Syntax

@findex GRANT
@findex REVOKE

@cindex privileges, granting
@cindex privileges, revoking
@cindex global privileges
@cindex revoking, privleges
@cindex granting, privleges

@example
GRANT priv_type [(column_list)] [, priv_type [(column_list)] ...]
    ON @{tbl_name | * | *.* | db_name.*@}
    TO user_name [IDENTIFIED BY [PASSWORD] 'password']
        [, user_name [IDENTIFIED BY 'password'] ...]
    [REQUIRE
    	[@{SSL| X509@}]
	[CIPHER cipher [AND]]
	[ISSUER issuer [AND]]
	[SUBJECT subject]]
    [WITH [GRANT OPTION | MAX_QUERIES_PER_HOUR=# |
                          MAX_UPDATES_PER_HOUR=# |
                          MAX_CONNECTIONS_PER_HOUR=#]]

REVOKE priv_type [(column_list)] [, priv_type [(column_list)] ...]
    ON @{tbl_name | * | *.* | db_name.*@}
    FROM user_name [, user_name ...]
@end example

@code{GRANT} is implemented in MySQL Version 3.22.11 or later. For
earlier MySQL versions, the @code{GRANT} statement does nothing.

The @code{GRANT} and @code{REVOKE} commands allow system administrators
to create users and grant and revoke rights to MySQL users at
four privilege levels:

@table @strong
@item Global level
Global privileges apply to all databases on a given server. These privileges
are stored in the @code{mysql.user} table.

@item Database level
Database privileges apply to all tables in a given database. These privileges
are stored in the @code{mysql.db} and @code{mysql.host} tables.

@item Table level
Table privileges apply to all columns in a given table. These privileges are
stored in the @code{mysql.tables_priv} table.

@item Column level
Column privileges apply to single columns in a given table. These privileges are
stored in the @code{mysql.columns_priv} table.
@end table

If you give a grant for a users that doesn't exists, that user is created.
For examples of how @code{GRANT} works, see @ref{Adding users}.

For the @code{GRANT} and @code{REVOKE} statements, @code{priv_type} may be
specified as any of the following:

@example
ALL PRIVILEGES      FILE                RELOAD
ALTER               INDEX               SELECT
CREATE              INSERT              SHUTDOWN
DELETE              PROCESS             UPDATE
DROP                REFERENCES          USAGE
@end example

@code{ALL} is a synonym for @code{ALL PRIVILEGES}.  @code{REFERENCES} is not
yet implemented.  @code{USAGE} is currently a synonym for ``no privileges.''
It can be used when you want to create a user that has no privileges.

To revoke the @strong{grant} privilege from a user, use a @code{priv_type}
value of @code{GRANT OPTION}:

@example
mysql> REVOKE GRANT OPTION ON ... FROM ...;
@end example

The only @code{priv_type} values you can specify for a table are @code{SELECT},
@code{INSERT}, @code{UPDATE}, @code{DELETE}, @code{CREATE}, @code{DROP},
@code{GRANT}, @code{INDEX}, and @code{ALTER}.

The only @code{priv_type} values you can specify for a column (that is, when
you use a @code{column_list} clause) are @code{SELECT}, @code{INSERT}, and
@code{UPDATE}.

You can set global privileges by using @code{ON *.*} syntax.  You can set
database privileges by using @code{ON db_name.*} syntax. If you specify
@code{ON *} and you have a current database, you will set the privileges for
that database.  (@strong{Warning}: if you specify @code{ON *} and you
@strong{don't} have a current database, you will affect the global privileges!)

In order to accommodate granting rights to users from arbitrary hosts,
MySQL supports specifying the @code{user_name} value in the form
@code{user@@host}.  If you want to specify a @code{user} string
containing special characters (such as @samp{-}), or a @code{host} string
containing special characters or wildcard characters (such as @samp{%}), you
can quote the user or host name (for example, @code{'test-user'@@'test-hostname'}).

You can specify wildcards in the hostname.  For example,
@code{user@@"%.loc.gov"} applies to @code{user} for any host in the
@code{loc.gov} domain, and @code{user@@"144.155.166.%"} applies to @code{user}
for any host in the @code{144.155.166} class C subnet.

The simple form @code{user} is a synonym for @code{user@@"%"}.

MySQL doesn't support wildcards in user names.  Anonymous users are
defined by inserting entries with @code{User=''} into the
@code{mysql.user} table or creating an user with an empty name with the
@code{GRANT} command.

@strong{Note}: if you allow anonymous users to connect to the MySQL
server, you should also grant privileges to all local users as
@code{user@@localhost} because otherwise the anonymous user entry for
the local host in the @code{mysql.user} table will be used when the user
tries to log into the MySQL server from the local machine!

You can verify if this applies to you by executing this query:

@example
mysql> SELECT Host,User FROM mysql.user WHERE User='';
@end example

For the moment, @code{GRANT} only supports host, table, database, and
column names up to 60 characters long. A user name can be up to 16
characters.

The privileges for a table or column are formed from the
logical OR of the privileges at each of the four privilege
levels.  For example, if the @code{mysql.user} table specifies that a
user has a global @strong{select} privilege, this can't be denied by an
entry at the database, table, or column level.

The privileges for a column can be calculated as follows:

@example
global privileges
OR (database privileges AND host privileges)
OR table privileges
OR column privileges
@end example

In most cases, you grant rights to a user at only one of the privilege
levels, so life isn't normally as complicated as above. The details of the
privilege-checking procedure are presented in
@ref{Privilege system}.

If you grant privileges for a user/hostname combination that does not exist
in the @code{mysql.user} table, an entry is added and remains there until
deleted with a @code{DELETE} command.  In other words, @code{GRANT} may
create @code{user} table entries, but @code{REVOKE} will not remove them;
you must do that explicitly using @code{DELETE}.

@cindex passwords, setting
In MySQL Version 3.22.12 or later,
if a new user is created or if you have global grant privileges, the user's
password will be set to the password specified by the @code{IDENTIFIED BY}
clause, if one is given.  If the user already had a password, it is replaced
by the new one.

If you don't want to send the password in clear text you can use the
@code{PASSWORD} option followed by a scrambled password from SQL
function @code{PASSWORD()} or the C API function
@code{make_scrambled_password(char *to, const char *password)}.

@strong{Warning}: if you create a new user but do not specify an
@code{IDENTIFIED BY} clause, the user has no password.  This is insecure.

Passwords can also be set with the @code{SET PASSWORD} command.
@xref{SET OPTION, , @code{SET OPTION}}.

If you grant privileges for a database, an entry in the @code{mysql.db}
table is created if needed. When all privileges for the database have been
removed with @code{REVOKE}, this entry is deleted.

If a user doesn't have any privileges on a table, the table is not displayed
when the user requests a list of tables (for example, with a @code{SHOW TABLES}
statement).

The @code{WITH GRANT OPTION} clause gives the user the ability to give
to other users any privileges the user has at the specified privilege level.
You should be careful to whom you give the @strong{grant} privilege, as two
users with different privileges may be able to join privileges!

@code{MAX_QUERIES_PER_HOUR=#}, @code{MAX_UPDATES_PER_HOUR=#} and
@code{MAX_CONNECTIONS_PER_HOUR=#} limit the number of
queries/updates and logins the user can do during one hour. 
If @code{#} is 0 (default), then this means that there is no limitations
for the user. @xref{User resources}.

You cannot grant another user a privilege you don't have yourself;
the @strong{grant} privilege allows you to give away only those privileges
you possess.

Be aware that when you grant a user the @strong{grant} privilege at a
particular privilege level, any privileges the user already possesses (or
is given in the future!) at that level are also grantable by that user.
Suppose you grant a user the @strong{insert} privilege on a database.  If
you then grant the @strong{select} privilege on the database and specify
@code{WITH GRANT OPTION}, the user can give away not only the @strong{select}
privilege, but also @strong{insert}.  If you then grant the @strong{update}
privilege to the user on the database, the user can give away the
@strong{insert}, @strong{select} and @strong{update}.

You should not grant @strong{alter} privileges to a normal user.  If you
do that, the user can try to subvert the privilege system by renaming
tables!

Note that if you are using table or column privileges for even one user, the
server examines table and column privileges for all users and this will slow
down MySQL a bit.

When @code{mysqld} starts, all privileges are read into memory.
Database, table, and column privileges take effect at once, and
user-level privileges take effect the next time the user connects.
Modifications to the grant tables that you perform using @code{GRANT} or
@code{REVOKE} are noticed by the server immediately.
If you modify the grant tables manually (using @code{INSERT}, @code{UPDATE},
etc.), you should execute a @code{FLUSH PRIVILEGES} statement or run
@code{mysqladmin flush-privileges} to tell the server to reload the grant
tables.
@xref{Privilege changes}.

@cindex ANSI SQL, differences from
The biggest differences between the ANSI SQL and MySQL versions of
@code{GRANT} are:

@itemize @bullet
@item
In MySQL privileges are given for an username + hostname combination
and not only for an username.

@item
ANSI SQL doesn't have global or database-level privileges, and ANSI SQL
doesn't support all privilege types that MySQL supports.
MySQL doesn't support the ANSI SQL @code{TRIGGER}, @code{EXECUTE} or
@code{UNDER} privileges.

@item
ANSI SQL privileges are structured in a hierarchal manner. If you remove
an user, all privileges the user has granted are revoked. In
MySQL the granted privileges are not automatically revoked, but
you have to revoke these yourself if needed.

@item
In MySQL, if you have the @strong{insert} privilege on only some of the
columns in a table, you can execute @code{INSERT} statements on the
table; the columns for which you don't have the @strong{insert} privilege
will be set to their default values. ANSI SQL requires you to have the
@strong{insert} privilege on all columns.

@item
When you drop a table in ANSI SQL, all privileges for the table are revoked.
If you revoke a privilege in ANSI SQL, all privileges that were granted based
on this privilege are also revoked. In MySQL, privileges can be
dropped only with explicit @code{REVOKE} commands or by manipulating the
MySQL grant tables.
@end itemize

For a description of using @code{REQUIRE}, see @ref{Secure connections}.

@node User names, Privilege changes, GRANT, User Account Management
@subsection MySQL User Names and Passwords

@cindex user names, and passwords
@cindex passwords, for users

There are several distinctions between the way user names and passwords are
used by MySQL and the way they are used by Unix or Windows:

@itemize @bullet
@item
User names, as used by MySQL for authentication purposes, have
nothing to do with Unix user names (login names) or Windows user names.  Most
MySQL clients by default try to log in using the current Unix user
name as the MySQL user name, but that is for convenience only.
Client programs allow a different name to be specified with the @code{-u} or
@code{--user} options. This means that you can't make a database secure in
any way unless all MySQL user names have passwords.  Anyone may
attempt to connect to the server using any name, and they will succeed if
they specify any name that doesn't have a password.

@item
MySQL user names can be up to 16 characters long; Unix user names
typically are limited to 8 characters.

@item
MySQL passwords have nothing to do with Unix passwords.  There is no
necessary connection between the password you use to log in to a Unix machine
and the password you use to access a database on that machine.

@item
MySQL encrypts passwords using a different algorithm than the
one used during the Unix login process.  See the descriptions of the
@code{PASSWORD()} and @code{ENCRYPT()} functions in @ref{Miscellaneous
functions}.  Note that even if the password is stored 'scrambled', and
knowing your 'scrambled' password is enough to be able to connect to
the MySQL server!
@end itemize

MySQL users and their privileges are normally created with the
@code{GRANT} command. @xref{GRANT}.

When you login to a MySQL server with a command-line client you
should specify the password with @code{--password=your-password}.
@xref{Connecting}.

@example
mysql --user=monty --password=guess database_name
@end example

If you want the client to prompt for a password, you should use
@code{--password} without any argument

@example
mysql --user=monty --password database_name
@end example

or the short form:

@example
mysql -u monty -p database_name
@end example

Note that in the last example the password is @strong{not} 'database_name'.

If you want to use the @code{-p} option to supply a password you should do so
like this:

@example
mysql -u monty -pguess database_name
@end example

On some systems, the library call that MySQL uses to prompt for a
password will automatically cut the password to 8 characters. Internally
MySQL doesn't have any limit for the length of the password.


@node Privilege changes, Default privileges, User names, User Account Management
@subsection When Privilege Changes Take Effect

When @code{mysqld} starts, all grant table contents are read into memory and
become effective at that point.

Modifications to the grant tables that you perform using @code{GRANT},
@code{REVOKE}, or @code{SET PASSWORD} are noticed by the server immediately.

If you modify the grant tables manually (using @code{INSERT}, @code{UPDATE},
etc.), you should execute a @code{FLUSH PRIVILEGES} statement or run
@code{mysqladmin flush-privileges} or @code{mysqladmin reload} to tell the
server to reload the grant tables. Otherwise, your changes will have @emph{no
effect} until you restart the server. If you change the grant tables manually
but forget to reload the privileges, you will be wondering why your changes
don't seem to make any difference!

When the server notices that the grant tables have been changed,
existing client connections are affected as follows:

@itemize @bullet
@item
Table and column privilege changes take effect with the client's next
request.

@item
Database privilege changes take effect at the next @code{USE db_name}
command.

@item
Global privilege changes and password changes take effect the next time
the client connects.
@end itemize


@node Default privileges, Adding users, Privilege changes, User Account Management
@subsection Setting Up the Initial MySQL Privileges

@cindex privileges, default
@cindex default, privileges
@cindex root password
@cindex superuser
@cindex users, root
@cindex anonymous user
@cindex password, root user

After installing MySQL, you set up the initial access privileges by
running @code{scripts/mysql_install_db}.
@xref{Quick install}.
The @code{mysql_install_db} script starts up the @code{mysqld}
server, then initialises the grant tables to contain the following set
of privileges:

@itemize @bullet
@item
The MySQL @code{root} user is created as a superuser who can do
anything.  Connections must be made from the local host.

@strong{Note}:
The initial @code{root} password is empty, so anyone can connect as @code{root}
@emph{without a password} and be granted all privileges.

@item
@cindex anonymous user
An anonymous user is created that can do anything with databases that have a
name of @code{'test'} or starting with @code{'test_'}.  Connections must be
made from the local host.  This means any local user can connect without a
password and be treated as the anonymous user.

@item
Other privileges are denied.  For example, normal users can't use
@code{mysqladmin shutdown} or @code{mysqladmin processlist}.
@end itemize

@strong{Note}: the default privileges are different for Windows.
@xref{Windows running}.

Because your installation is initially wide open, one of the first things you
should do is specify a password for the MySQL
@code{root} user.  You can do this as follows (note that you specify the
password using the @code{PASSWORD()} function):

@example
shell> mysql -u root mysql
mysql> SET PASSWORD FOR root@@localhost=PASSWORD('new_password');
@end example

If you know what you are doing, you can also directly manipulate the
privilege tables:

@example
shell> mysql -u root mysql
mysql> UPDATE user SET Password=PASSWORD('new_password')
    ->     WHERE user='root';
mysql> FLUSH PRIVILEGES;
@end example

Another way to set the password is by using the @code{mysqladmin} command:

@example
shell> mysqladmin -u root password new_password
@end example

Only users with write/update access to the @code{mysql} database can change the
password for others users.  All normal users (not anonymous ones) can only
change their own password with either of the above commands or with
@code{SET PASSWORD=PASSWORD('new password')}.

Note that if you update the password in the @code{user} table directly using
the first method, you must tell the server to re-read the grant tables (with
@code{FLUSH PRIVILEGES}), because the change will go unnoticed otherwise.

Once the @code{root} password has been set, thereafter you must supply that
password when you connect to the server as @code{root}.

You may wish to leave the @code{root} password blank so that you don't need
to specify it while you perform additional setup or testing. However, be sure
to set it before using your installation for any real production work.

See the @code{scripts/mysql_install_db} script to see how it sets up
the default privileges.  You can use this as a basis to see how to
add other users.

If you want the initial privileges to be different from those just described
above, you can modify @code{mysql_install_db} before you run it.

@cindex grant tables, re-creating
@cindex re-creating, grant tables
To re-create the grant tables completely, remove all the @file{.frm},
@file{.MYI}, and @file{.MYD} files in the directory containing the
@code{mysql} database.  (This is the directory named @file{mysql} under
the database directory, which is listed when you run @code{mysqld
--help}.) Then run the @code{mysql_install_db} script, possibly after
editing it first to have the privileges you want.

@strong{Note}: for MySQL versions older than Version 3.22.10,
you should not delete the @file{.frm} files.  If you accidentally do this,
you should copy them back from your MySQL distribution before
running @code{mysql_install_db}.


@node Adding users, User resources, Default privileges, User Account Management
@subsection Adding New Users to MySQL

@findex GRANT statement
@findex statements, GRANT

@cindex privileges, adding
@cindex adding, new user privileges
@cindex user privileges, adding

You can add users two different ways: by using @code{GRANT} statements
or by manipulating the MySQL grant tables directly.  The
preferred method is to use @code{GRANT} statements, because they are
more concise and less error-prone. @xref{GRANT}.

There are also a lot of contributed programs like @code{phpmyadmin}
that can be used to create and administrate users. @xref{Portals}.

The following examples show how to use the @code{mysql} client to set up new
users.  These examples assume that privileges are set up according to the
defaults described in the previous section.  This means that to make changes,
you must be on the same machine where @code{mysqld} is running, you must
connect as the MySQL @code{root} user, and the @code{root} user must
have the @strong{insert} privilege for the @code{mysql} database and the
@strong{reload} administrative privilege.  Also, if you have changed the
@code{root} user password, you must specify it for the @code{mysql} commands here.

You can add new users by issuing @code{GRANT} statements:

@example
shell> mysql --user=root mysql
mysql> GRANT ALL PRIVILEGES ON *.* TO monty@@localhost
    ->     IDENTIFIED BY 'some_pass' WITH GRANT OPTION;
mysql> GRANT ALL PRIVILEGES ON *.* TO monty@@"%"
    ->     IDENTIFIED BY 'some_pass' WITH GRANT OPTION;
mysql> GRANT RELOAD,PROCESS ON *.* TO admin@@localhost;
mysql> GRANT USAGE ON *.* TO dummy@@localhost;
@end example

These @code{GRANT} statements set up three new users:

@table @code
@item monty
A full superuser who can connect to the server from anywhere, but who must
use a password @code{'some_pass'} to do so.  Note that we must issue
@code{GRANT} statements for both @code{monty@@localhost} and
@code{monty@@"%"}.  If we don't add the entry with @code{localhost}, the
anonymous user entry for @code{localhost} that is created by
@code{mysql_install_db} will take precedence when we connect from the local
host, because it has a more specific @code{Host} field value and thus comes
earlier in the @code{user} table sort order.

@item admin
A user who can connect from @code{localhost} without a password and who is
granted the @strong{reload} and @strong{process} administrative privileges.
This allows the user to execute the @code{mysqladmin reload},
@code{mysqladmin refresh}, and @code{mysqladmin flush-*} commands, as well as
@code{mysqladmin processlist} .  No database-related privileges are granted.
(They can be granted later by issuing additional @code{GRANT} statements.)

@item dummy
A user who can connect without a password, but only from the local host.  The
global privileges are all set to @code{'N'}@-the @strong{usage} privilege
type allows you to create a user with no privileges.  It is assumed that you
will grant database-specific privileges later.
@end table

@findex INSERT statement, grant privileges
@findex statements, INSERT
You can also add the same user access information directly by issuing
@code{INSERT} statements and then telling the server to reload the grant
tables:

@example
shell> mysql --user=root mysql
mysql> INSERT INTO user VALUES('localhost','monty',PASSWORD('some_pass'),
    ->          'Y','Y','Y','Y','Y','Y','Y','Y','Y','Y','Y','Y','Y','Y');
mysql> INSERT INTO user VALUES('%','monty',PASSWORD('some_pass'),
    ->          'Y','Y','Y','Y','Y','Y','Y','Y','Y','Y','Y','Y','Y','Y');
mysql> INSERT INTO user SET Host='localhost',User='admin',
    ->           Reload_priv='Y', Process_priv='Y';
mysql> INSERT INTO user (Host,User,Password)
    ->                  VALUES('localhost','dummy','');
mysql> FLUSH PRIVILEGES;
@end example

Depending on your MySQL version, you may have to use a different
number of @code{'Y'} values above (versions prior to Version 3.22.11 had fewer
privilege columns).  For the @code{admin} user, the more readable extended
@code{INSERT} syntax that is available starting with Version 3.22.11 is used.

Note that to set up a superuser, you need only create a @code{user} table
entry with the privilege fields set to @code{'Y'}.  No @code{db} or
@code{host} table entries are necessary.

The privilege columns in the @code{user} table were not set explicitly in the
last @code{INSERT} statement (for the @code{dummy} user), so those columns
are assigned the default value of @code{'N'}.  This is the same thing that
@code{GRANT USAGE} does.

The following example adds a user @code{custom} who can connect from hosts
@code{localhost}, @code{server.domain}, and @code{whitehouse.gov}.  He wants
to access the @code{bankaccount} database only from @code{localhost},
the @code{expenses} database only from @code{whitehouse.gov}, and
the @code{customer} database from all three hosts.  He wants
to use the password @code{stupid} from all three hosts.

To set up this user's privileges using @code{GRANT} statements, run these
commands:

@example
shell> mysql --user=root mysql
mysql> GRANT SELECT,INSERT,UPDATE,DELETE,CREATE,DROP
    ->     ON bankaccount.*
    ->     TO custom@@localhost
    ->     IDENTIFIED BY 'stupid';
mysql> GRANT SELECT,INSERT,UPDATE,DELETE,CREATE,DROP
    ->     ON expenses.*
    ->     TO custom@@whitehouse.gov
    ->     IDENTIFIED BY 'stupid';
mysql> GRANT SELECT,INSERT,UPDATE,DELETE,CREATE,DROP
    ->     ON customer.*
    ->     TO custom@@'%'
    ->     IDENTIFIED BY 'stupid';
@end example

The reason that we do to grant statements for the user 'custom' is that
we want the give the user access to MySQL both from the local
machine with Unix sockets and from the remote machine 'whitehouse.gov'
over TCP/IP.

To set up the user's privileges by modifying the grant tables directly,
run these commands (note the @code{FLUSH PRIVILEGES} at the end):

@example
shell> mysql --user=root mysql
mysql> INSERT INTO user (Host,User,Password)
    -> VALUES('localhost','custom',PASSWORD('stupid'));
mysql> INSERT INTO user (Host,User,Password)
    -> VALUES('server.domain','custom',PASSWORD('stupid'));
mysql> INSERT INTO user (Host,User,Password)
    -> VALUES('whitehouse.gov','custom',PASSWORD('stupid'));
mysql> INSERT INTO db
    -> (Host,Db,User,Select_priv,Insert_priv,Update_priv,Delete_priv,
    ->  Create_priv,Drop_priv)
    -> VALUES
    -> ('localhost','bankaccount','custom','Y','Y','Y','Y','Y','Y');
mysql> INSERT INTO db
    -> (Host,Db,User,Select_priv,Insert_priv,Update_priv,Delete_priv,
    ->  Create_priv,Drop_priv)
    -> VALUES
    -> ('whitehouse.gov','expenses','custom','Y','Y','Y','Y','Y','Y');
mysql> INSERT INTO db
    -> (Host,Db,User,Select_priv,Insert_priv,Update_priv,Delete_priv,
    ->  Create_priv,Drop_priv)
    -> VALUES('%','customer','custom','Y','Y','Y','Y','Y','Y');
mysql> FLUSH PRIVILEGES;
@end example

The first three @code{INSERT} statements add @code{user} table entries that
allow user @code{custom} to connect from the various hosts with the given
password, but grant no permissions to him (all privileges are set to the
default value of @code{'N'}).  The next three @code{INSERT} statements add
@code{db} table entries that grant privileges to @code{custom} for the
@code{bankaccount}, @code{expenses}, and @code{customer} databases, but only
when accessed from the proper hosts.  As usual, when the grant tables are
modified directly, the server must be told to reload them (with
@code{FLUSH PRIVILEGES}) so that the privilege changes take effect.

If you want to give a specific user access from any machine in a given
domain, you can issue a @code{GRANT} statement like the following:

@example
mysql> GRANT ...
    ->     ON *.*
    ->     TO myusername@@"%.mydomainname.com"
    ->     IDENTIFIED BY 'mypassword';
@end example

To do the same thing by modifying the grant tables directly, do this:

@example
mysql> INSERT INTO user VALUES ('%.mydomainname.com', 'myusername',
    ->             PASSWORD('mypassword'),...);
mysql> FLUSH PRIVILEGES;
@end example

You can also use @code{xmysqladmin}, @code{mysql_webadmin}, and even
@code{xmysql} to insert, change, and update values in the grant tables.
You can find these utilities in the Contrib directory of the
MySQL web site (@uref{http://www.mysql.com/Downloads/Contrib/}).

@node User resources, Passwords, Adding users, User Account Management
@subsection Limiting user resources

Starting from MySQL 4.0.2 one can limit certain resources per user.

So far, the only available method of limiting usage of MySQL
server resources has been setting the @code{max_user_connections}
startup variable to a non-zero value. But this method is strictly
global and does not allow for management of individual users, which
could be of particular interest to Internet Service Providers.

Therefore, management of three resources is introduced on the
individual user level:

@itemize @bullet
@item Number of all queries per hour:
All commands that could be run by a user.
@item Number of all updates per hour:
Any command that changes any table or database.
@item Number of connections made per hour:
New connections opened per hour.
@end itemize

A user in the aforementioned context is a single entry in the
@code{user} table, which is uniquely identified by its @code{user}
and @code{host} columns.

All users are by default not limited in using the above resources,
unless the limits are granted to them. These limits can be granted
@strong{only} via global @code{GRANT (*.*)}, using this syntax:

@example
GRANT ... WITH MAX_QUERIES_PER_HOUR = N1
               MAX_UPDATES_PER_HOUR = N2
               MAX_CONNECTIONS_PER_HOUR = N3;
@end example

One can specify any combination of the above resources.
N1, N2 and N3 are integers and stands for count / hour.

If user reaches any of the above limits withing one hour, his
connection will be terminated or refused and the appropriate error
message shall be issued.

Current usage values for a particular user can be flushed (set to zero)
by issuing a @code{GRANT} statement with any of the above clauses,
including a @code{GRANT} statement with the current values.

Also, current values for all users will be flushed if privileges are
reloaded (in the server or using @code{mysqladmin reload})
or if the @code{FLUSH USER_RESOURCES} command is issued.

The feature is enabled as soon as a single user is granted with any
of the limiting @code{GRANT} clauses.

As a prerequisite for enabling this feature, the @code{user} table in
the @code{mysql} database must contain the additional columns, as
defined in the table creation scripts @code{mysql_install_db} and
@code{mysql_install_db.sh} in @file{scripts} subdirectory.

@node Passwords, Password security, User resources, User Account Management
@subsection Setting Up Passwords

@findex PASSWORD()
@findex SET PASSWORD statement

@cindex passwords, setting
@cindex setting, passwords

In most cases you should use @code{GRANT} to set up your users/passwords,
so the following only applies for advanced users. @xref{GRANT, , @code{GRANT}}.

The examples in the preceding sections illustrate an important principle:
when you store a non-empty password using @code{INSERT} or @code{UPDATE}
statements, you must use the @code{PASSWORD()} function to encrypt it.  This
is because the @code{user} table stores passwords in encrypted form, not as
plaintext.  If you forget that fact, you are likely to attempt to set
passwords like this:

@example
shell> mysql -u root mysql
mysql> INSERT INTO user (Host,User,Password)
    -> VALUES('%','jeffrey','biscuit');
mysql> FLUSH PRIVILEGES;
@end example

The result is that the plaintext value @code{'biscuit'} is stored as the
password in the @code{user} table.  When the user @code{jeffrey} attempts to
connect to the server using this password, the @code{mysql} client encrypts
it with @code{PASSWORD()}, generates an authentification vector
based on @strong{encrypted} password and a random number,
obtained from server, and sends the result to the server.
The server uses the @code{password} value in the @code{user} table
(that is @strong{not encrypted} value @code{'biscuit'})
to perform the same calculations, and compares results.
The comparison fails and the server rejects the
connection:

@example
shell> mysql -u jeffrey -pbiscuit test
Access denied
@end example

Passwords must be encrypted when they are inserted in the @code{user}
table, so the @code{INSERT} statement should have been specified like this
instead:

@example
mysql> INSERT INTO user (Host,User,Password)
    -> VALUES('%','jeffrey',PASSWORD('biscuit'));
@end example

You must also use the @code{PASSWORD()} function when you use @code{SET
PASSWORD} statements:

@example
mysql> SET PASSWORD FOR jeffrey@@"%" = PASSWORD('biscuit');
@end example

If you set passwords using the @code{GRANT ... IDENTIFIED BY} statement
or the @code{mysqladmin password} command, the @code{PASSWORD()} function
is unnecessary.  They both take care of encrypting the password for you,
so you would specify a password of @code{'biscuit'} like this:

@example
mysql> GRANT USAGE ON *.* TO jeffrey@@"%" IDENTIFIED BY 'biscuit';
@end example

or

@example
shell> mysqladmin -u jeffrey password biscuit
@end example

@strong{Note}: @code{PASSWORD()} does not perform password encryption in the
same way that Unix passwords are encrypted.  You should not assume that if
your Unix password and your MySQL password are the same, that
@code{PASSWORD()} will result in the same encrypted value as is stored in the
Unix password file. @xref{User names}.


@node Password security, Secure connections, Passwords, User Account Management
@subsection Keeping Your Password Secure

It is inadvisable to specify your password in a way that exposes it to
discovery by other users.  The methods you can use to specify your password
when you run client programs are listed here, along with an assessment of
the risks of each method:

@itemize @bullet
@item
Never give a normal user access to the @code{mysql.user} table. Knowing
the encrypted password for a user makes it possible to login as this
user.  The passwords are only scrambled so that one shouldn't be able to
see the real password you used (if you happen to use a similar password
with your other applications).

@item
Use a @code{-pyour_pass} or @code{--password=your_pass} option on the command
line.  This is convenient but insecure, because your password becomes visible
to system status programs (such as @code{ps}) that may be invoked by other
users to display command-lines.  (MySQL clients typically overwrite
the command-line argument with zeroes during their initialisation sequence,
but there is still a brief interval during which the value is visible.)

@item
Use a @code{-p} or @code{--password} option (with no @code{your_pass} value
specified).  In this case, the client program solicits the password from
the terminal:
@findex -p option
@findex -password option

@example
shell> mysql -u user_name -p
Enter password: ********
@end example

The @samp{*} characters represent your password.

It is more secure to enter your password this way than to specify it on the
command-line because it is not visible to other users.  However, this method
of entering a password is suitable only for programs that you run
interactively.  If you want to invoke a client from a script that runs
non-interactively, there is no opportunity to enter the password from the
terminal. On some systems, you may even find that the first line of your
script is read and interpreted (incorrectly) as your password!

@item
@tindex .my.cnf file
Store your password in a configuration file.  For example, you can list your
password in the @code{[client]} section of the @file{.my.cnf} file in your
home directory:

@example
[client]
password=your_pass
@end example

If you store your password in @file{.my.cnf}, the file should not be group or
world readable or writable.  Make sure the file's access mode is @code{400}
or @code{600}.

@xref{Option files}.
@item
You can store your password in the @code{MYSQL_PWD} environment variable, but
this method must be considered extremely insecure and should not be used.
Some versions of @code{ps} include an option to display the environment of
running processes; your password will be in plain sight for all to see if
you set @code{MYSQL_PWD}.  Even on systems without such a version of
@code{ps}, it is unwise to assume there is no other method to observe process
environments. @xref{Environment variables}.
@end itemize

All in all, the safest methods are to have the client program prompt for the
password or to specify the password in a properly protected @file{.my.cnf}
file.

@node Secure connections,  , Password security, User Account Management
@subsection Using Secure Connections

@cindex openssl
@cindex SSL and X509 Basics

@menu
* Secure basics::               Basics
* Secure requirements::         Requirements
* Secure GRANT::                @code{GRANT} Options
@end menu

@node Secure basics, Secure requirements, Secure connections, Secure connections
@subsubsection Basics

MySQL has support for SSL encrypted connections. To understand how MySQL
uses SSL, we need to explain some basics about SSL and X509. People who
are already aware of it can skip this part.

By default, MySQL uses unencrypted connections between client and
server. This means that someone could watch all your traffic and look at
the data being sent/received.  Actually, they could even change the data
while it is in transit between client and server. Sometimes you need to
move really secret data over public networks and in such a case using an
unencrypted connection is unacceptable.

SSL is a protocol which uses different encryption algorithms to ensure
that data which comes from public network can be trusted. It has
mechanisms to detect any change, loss or replay of data. SSL also
incorpores algorithms to recognise and provide identity verification
using the X509 standard.

@cindex What is encryption
Encryption is the way to make any kind of data unreadable. In fact,
today's practice requires many additional security elements from
encryption algorithms.  They should resist many kind of known attacks
like just messing with order of encrypted messages or replaying data
twice.

@cindex What is X509/Certificate?
X509 is a standard that makes it possible to identify someone in the
Internet.  It is most commonly used in e-commerce applications. In basic
terms, there should be some company called "Certificate Authority" which
assigns electronic certificates to anyone who needs them. Certificates
rely on asymmetric encryption algorithms which have two encryption keys
- public and secret. A certificate owner can prove his identity by
showing his certificate to other party. A certificate consists of his
owner's public key. Any data encrypted with this public key can only be
decrypted using the corresponding secret key, which is held by the owner
of the certificate.

MySQL doesn't use encrypted on connections by default, because this
would make the client/server protocol much slower. Any kind of
additional functionality requires computer to do additional work and
encrypting data is CPU-intensive operation require time and can delay
MySQL main tasks. By default MySQL is tuned to be fast as possible.

If you need more information about SSL/X509/encryption, you should use
your favourite internet search engine and search for keywords you are
interested in.

@node Secure requirements, Secure GRANT, Secure basics, Secure connections
@subsubsection Requirements

To get secure connections to work with MySQL you must do the following:

@enumerate
@item
Install the openssl library. We have tested MySQL with openssl 0.9.6.
@uref{http://www.openssl.org/}.
@item
Configure MySQL with @code{--with-vio --with-openssl}.
@item
If you are using an old MySQL installation, you have to update your
@code{mysql.user} table with some new columns. You can do this by
running the @code{mysql_fix_privilege_tables.sh} script.
@item
You can check if a running mysqld server supports @code{openssl} by
examining if @code{SHOW VARIABLES LIKE 'have_openssl'} returns @code{YES}.
@end enumerate


@node Secure GRANT,  , Secure requirements, Secure connections
@subsubsection @code{GRANT} Options

@cindex SSL related options
@findex REQUIRE GRANT option
@findex GRANT statemenet

MySQL can check X509 certificate attributes in addition to the
normal username/password scheme. All the usual options are still
required (username, password, IP address mask, database/table name).

There are different possibilities to limit connections:

@itemize @bullet
@item
Without any SSL/X509 options, all kind of encrypted/unencrypted
connections are allowed if username and password are valid.

@item
@code{REQUIRE SSL} option limits the server to allow only SSL
encrypted connections. Note that this option can be omitted
if there are any ACL records which allow non-SSL connections.

@example
mysql> GRANT ALL PRIVILEGES ON test.* TO root@@localhost
    -> IDENTIFIED BY "goodsecret" REQUIRE SSL;
@end example

@item
@code{REQUIRE X509} means that client should have valid certificate
but we do not care about the exact certificate, issuer or subject.
The only restriction is that it should be possible to verify its
signature with one of the CA certificates.

@example
mysql> GRANT ALL PRIVILEGES ON test.* TO root@@localhost
    -> IDENTIFIED BY "goodsecret" REQUIRE X509;
@end example

@item
@code{REQUIRE ISSUER issuer} makes connection more restrictive: now
client must present a valid X509 certificate issued by CA "issuer".
Using X509 certificates always implies encryption, so the option "SSL"
is not neccessary anymore.

@example
mysql> GRANT ALL PRIVILEGES ON test.* TO root@@localhost
    -> IDENTIFIED BY "goodsecret"
    -> REQUIRE ISSUER "C=FI, ST=Some-State, L=Helsinki,
    "> O=MySQL Finland AB, CN=Tonu Samuel/Email=tonu@@mysql.com";
@end example

@item
@code{REQUIRE SUBJECT subject} requires clients to have valid X509
certificate with subject "subject" on it. If client have valid
certificate but having different "subject" then the connection is
still not allowed.

@example
mysql> GRANT ALL PRIVILEGES ON test.* TO root@@localhost
    -> IDENTIFIED BY "goodsecret"
    -> REQUIRE SUBJECT "C=EE, ST=Some-State, L=Tallinn,
    "> O=MySQL demo client certificate,
    "> CN=Tonu Samuel/Email=tonu@@mysql.com";
@end example

@item
@code{REQUIRE CIPHER cipher} is needed to assure enough strong ciphers
and keylengths will be used. SSL itself can be weak if old algorithms
with short encryption keys are used. Using this option, we can ask for
some exact cipher method to allow a connection.

@example
mysql> GRANT ALL PRIVILEGES ON test.* TO root@@localhost
    -> IDENTIFIED BY "goodsecret"
    -> REQUIRE CIPHER "EDH-RSA-DES-CBC3-SHA";
@end example

Also it is allowed to combine these options with each other like this:

@example
mysql> GRANT ALL PRIVILEGES ON test.* TO root@@localhost
    -> IDENTIFIED BY "goodsecret"
    -> REQUIRE SUBJECT "C=EE, ST=Some-State, L=Tallinn,
    "> O=MySQL demo client certificate,
    "> CN=Tonu Samuel/Email=tonu@@mysql.com"
    -> AND ISSUER "C=FI, ST=Some-State, L=Helsinki,
    "> O=MySQL Finland AB, CN=Tonu Samuel/Email=tonu@@mysql.com"
    -> AND CIPHER "EDH-RSA-DES-CBC3-SHA";
@end example

But it is not allowed to use any option twice. Only different
options can be mixed.
@end itemize


@node Disaster Prevention, Database Administration, User Account Management, MySQL Database Administration
@section Disaster Prevention and Recovery

@c FIX: add section intro

@menu
* Backup::                      Database Backups
* BACKUP TABLE::                @code{BACKUP TABLE} Syntax
* RESTORE TABLE::               @code{RESTORE TABLE} Syntax
* CHECK TABLE::                 @code{CHECK TABLE} Syntax
* REPAIR TABLE::                @code{REPAIR TABLE} Syntax
* Table maintenance::           Using @code{myisamchk} for Table Maintenance and Crash Recovery
* Maintenance regimen::         Setting Up a Table Maintenance Regimen
* Table-info::                  Getting Information About a Table
@end menu


@node Backup, BACKUP TABLE, Disaster Prevention, Disaster Prevention
@subsection Database Backups

@cindex databases, backups
@cindex backups

Because MySQL tables are stored as files, it is easy to do a
backup. To get a consistent backup, do a @code{LOCK TABLES} on the
relevant tables followed by @code{FLUSH TABLES} for the tables.
@xref{LOCK TABLES, , @code{LOCK TABLES}}.
@xref{FLUSH, , @code{FLUSH}}.
You only need a read lock; this allows other threads to continue to
query the tables while you are making a copy of the files in the
database directory.  The @code{FLUSH TABLE} is needed to ensure that
the all active index pages is written to disk before you start the backup.

If you want to make a SQL level backup of a table, you can use
@code{SELECT INTO OUTFILE} or @code{BACKUP TABLE}. @xref{SELECT}.
@xref{BACKUP TABLE}.

Another way to back up a database is to use the @code{mysqldump} program or
the @code{mysqlhotcopy script}. @xref{mysqldump, , @code{mysqldump}}.
@xref{mysqlhotcopy, , @code{mysqlhotcopy}}.

@enumerate
@item
Do a full backup of your databases:

@example
shell> mysqldump --tab=/path/to/some/dir --opt --full

or

shell> mysqlhotcopy database /path/to/some/dir
@end example

You can also simply copy all table files (@file{*.frm}, @file{*.MYD}, and
@file{*.MYI} files) as long as the server isn't updating anything.
The script @code{mysqlhotcopy} does use this method.

@item
@cindex log files, names
Stop @code{mysqld} if it's running, then start it with the
@code{--log-update[=file_name]} option.  @xref{Update log}. The update
log file(s) provide you with the information you need to replicate
changes to the database that are made subsequent to the point at which
you executed @code{mysqldump}.
@end enumerate

If you have to restore something, try to recover your tables using
@code{REPAIR TABLE} or @code{myisamchk -r} first.  That should work in
99.9% of all cases.  If @code{myisamchk} fails, try the following
procedure (this will only work if you have started MySQL with
@code{--log-update}, @pxref{Update log}):

@enumerate
@item
Restore the original @code{mysqldump} backup.
@item
Execute the following command to re-run the updates in the binary log:

@example
shell> mysqlbinlog hostname-bin.[0-9]* | mysql
@end example

If you are using the update log you can use:

@example
shell> ls -1 -t -r hostname.[0-9]* | xargs cat | mysql
@end example
@end enumerate

@code{ls} is used to get all the update log files in the right order.

You can also do selective backups with @code{SELECT * INTO OUTFILE 'file_name'
FROM tbl_name} and restore with @code{LOAD DATA INFILE 'file_name' REPLACE
...} To avoid duplicate records, you need a @code{PRIMARY KEY} or a
@code{UNIQUE} key in the table. The @code{REPLACE} keyword causes old records
to be replaced with new ones when a new record duplicates an old record on
a unique key value.

If you get performance problems in making backups on your system, you can
solve this by setting up replication and do the backups on the slave
instead of on the master. @xref{Replication Intro}.

If you are using a Veritas filesystem, you can do:

@enumerate
@item
From a client (or Perl), execute: @code{FLUSH TABLES WITH READ LOCK}.

@item
From another shell, execute: @code{mount vxfs snapshot}.

@item
From the first client, execute: @code{UNLOCK TABLES}.

@item
Copy files from snapshot.

@item
Unmount snapshot.
@end enumerate


@node BACKUP TABLE, RESTORE TABLE, Backup, Disaster Prevention
@subsection @code{BACKUP TABLE} Syntax

@findex BACKUP TABLE

@cindex backups, database

@example
BACKUP TABLE tbl_name[,tbl_name...] TO '/path/to/backup/directory'
@end example

Copies to the backup directory the minimum number of table files needed
to restore the table. Currently only works for @code{MyISAM} tables.
For @code{MyISAM} tables, copies @file{.frm} (definition)  and
@file{.MYD} (data) files. The index file can be rebuilt from those two.

Before using this command, please see @ref{Backup}.

During the backup, read lock will be held for each table, one at time,
as they are being backed up. If you want to backup several tables as
a snapshot, you must first issue @code{LOCK TABLES} obtaining a read
lock for each table in the group.

The command returns a table with the following columns:

@multitable @columnfractions .15 .40
@item @strong{Column} @tab @strong{Value}
@item Table @tab Table name
@item Op @tab Always ``backup''
@item Msg_type @tab One of @code{status}, @code{error}, @code{info} or @code{warning}.
@item Msg_text @tab The message.
@end multitable

Note that @code{BACKUP TABLE} is only available in MySQL
version 3.23.25 and later.


@node RESTORE TABLE, CHECK TABLE, BACKUP TABLE, Disaster Prevention
@subsection @code{RESTORE TABLE} Syntax

@findex RESTORE TABLE

@example
RESTORE TABLE tbl_name[,tbl_name...] FROM '/path/to/backup/directory'
@end example

Restores the table(s) from the backup that was made with
@code{BACKUP TABLE}. Existing tables will not be overwritten - if you
try to restore over an existing table, you will get an error. Restore
will take  longer than BACKUP due to the need to rebuilt the index. The
more keys you have, the longer it is going to take. Just as
@code{BACKUP TABLE}, currently only works of @code{MyISAM} tables.


The command returns a table with the following columns:

@multitable @columnfractions .15 .40
@item @strong{Column} @tab @strong{Value}
@item Table @tab Table name
@item Op @tab Always ``restore''
@item Msg_type @tab One of @code{status}, @code{error}, @code{info} or @code{warning}.
@item Msg_text @tab The message.
@end multitable


@node CHECK TABLE, REPAIR TABLE, RESTORE TABLE, Disaster Prevention
@subsection @code{CHECK TABLE} Syntax

@findex CHECK TABLE

@example
CHECK TABLE tbl_name[,tbl_name...] [option [option...]]

option = QUICK | FAST | MEDIUM | EXTENDED | CHANGED
@end example

@code{CHECK TABLE} only works on @code{MyISAM} and @code{InnoDB} tables. On
@code{MyISAM} tables it's the same thing as running @code{myisamchk -m
table_name} on the table.

If you don't specify any option @code{MEDIUM} is used.

Checks the table(s) for errors. For @code{MyISAM} tables the key statistics
is updated. The command returns a table with the following columns:

@multitable @columnfractions .15 .40
@item @strong{Column} @tab @strong{Value}
@item Table    @tab Table name.
@item Op       @tab Always ``check''.
@item Msg_type @tab One of @code{status}, @code{error}, @code{info}, or @code{warning}.
@item Msg_text @tab The message.
@end multitable

Note that you can get many rows of information for each checked
table. The last row will be of @code{Msg_type status} and should
normally be @code{OK}.  If you don't get @code{OK}, or @code{Not
checked} you should normally run a repair of the table. @xref{Table
maintenance}. @code{Not checked} means that the table the given @code{TYPE}
told MySQL that there wasn't any need to check the table.

The different check types stand for the following:

@multitable @columnfractions .15 .85
@item @strong{Type} @tab @strong{Meaning}
@item @code{QUICK} @tab Don't scan the rows to check for wrong links.
@item @code{FAST}  @tab Only check tables which haven't been closed properly.
@item @code{CHANGED} @tab Only check tables which have been changed since last check or haven't been closed properly.
@item @code{MEDIUM} @tab Scan rows to verify that deleted links are okay. This also calculates a key checksum for the rows and verifies this with a calcualted checksum for the keys.
@item @code{EXTENDED} @tab Do a full key lookup for all keys for each row.  This ensures that the table is 100 % consistent, but will take a long time!
@end multitable

For dynamically sized @code{MyISAM} tables a started check will always
do a @code{MEDIUM} check. For static size rows we skip the row scan
for @code{QUICK} and @code{FAST} as the rows are very seldom corrupted.

You can combine check options as in:

@example
CHECK TABLE test_table FAST QUICK;
@end example

Which would simply do a quick check on the table to see whether it was
closed properly.

@strong{Note}: that in some case @code{CHECK TABLE} will change the
table!  This happens if the table is marked as 'corrupted' or 'not
closed properly' but @code{CHECK TABLE} didn't find any problems in the
table.  In this case @code{CHECK TABLE} will mark the table as okay.

If a table is corrupted, then it's most likely that the problem is in
the indexes and not in the data part.  All of the above check types
checks the indexes thoroughly and should thus find most errors.

If you just want to check a table that you assume is okay, you should use
no check options or the @code{QUICK} option. The latter should be used
when you are in a hurry and can take the very small risk that
@code{QUICK} didn't find an error in the datafile. (In most cases
MySQL should find, under normal usage, any error in the data
file. If this happens then the table will be marked as 'corrupted',
in which case the table can't be used until it's repaired.)

@code{FAST} and @code{CHANGED} are mostly intended to be used from a
script (for example to be executed from cron) if you want to check your
table from time to time. In most cases you @code{FAST} is to be prefered
over @code{CHANGED}.  (The only case when it isn't is when you suspect a
bug you have found a bug in the @code{MyISAM} code.)

@code{EXTENDED} is only to be used after you have run a normal check but
still get strange errors from a table when MySQL tries to
update a row or find a row by key (this is very unlikely if a
normal check has succeeded!).

Some things reported by check table, can't be corrected automatically:

@itemize @bullet
@item
@code{Found row where the auto_increment column has the value 0}.

This means that you have in the table a row where the
@code{AUTO_INCREMENT} index column contains the value 0.
(It's possible to create a row where the @code{AUTO_INCREMENT} column is 0 by
explicitly setting the column to 0 with an @code{UPDATE} statement)

This isn't an error in itself, but could cause trouble if you decide to
dump the table and restore it or do an @code{ALTER TABLE} on the
table. In this case the @code{AUTO_INCREMENT} column will change value,
according to the rules of @code{AUTO_INCREMENT} columns, which could cause
problems like a duplicate key error.

To get rid of the warning, just execute an @code{UPDATE} statement
to set the column to some other value than 0.
@end itemize


@node REPAIR TABLE, Table maintenance, CHECK TABLE, Disaster Prevention
@subsection @code{REPAIR TABLE} Syntax

@findex REPAIR TABLE

@example
REPAIR TABLE tbl_name[,tbl_name...] [QUICK] [EXTENDED] [USE_FRM]
@end example

@code{REPAIR TABLE} only works on @code{MyISAM} tables and is the same
as running @code{myisamchk -r table_name} on the table.

Normally you should never have to run this command, but if disaster strikes
you are very likely to get back all your data from a MyISAM table with
@code{REPAIR TABLE}. If your tables get corrupted a lot you should
try to find the reason for this! @xref{Crashing}. @xref{MyISAM table problems}.

@code{REPAIR TABLE} repairs a possible corrupted table. The command returns a
table with the following columns:

@multitable @columnfractions .15 .40
@item @strong{Column} @tab @strong{Value}
@item Table @tab Table name
@item Op @tab Always ``repair''
@item Msg_type @tab One of @code{status}, @code{error}, @code{info} or @code{warning}.
@item Msg_text @tab The message.
@end multitable

Note that you can get many rows of information for each repaired
table. The last one row will be of @code{Msg_type status} and should
normally be @code{OK}.  If you don't get @code{OK}, you should try
repairing the table with @code{myisamchk -o}, as @code{REPAIR TABLE}
does not yet implement all the options of @code{myisamchk}. In the near
future, we will make it more flexible.

If @code{QUICK} is given then MySQL will try to do a
@code{REPAIR} of only the index tree.

If you use @code{EXTENDED} then MySQL will create the index row
by row instead of creating one index at a time with sorting; this may be
better than sorting on fixed-length keys if you have long @code{char()}
keys that compress very good.

As of @code{MySQL} 4.0.2 there is a @code{USE_FRM} mode for @code{REPAIR}.
Use it if the @file{.MYI} file is missing or if its header is corrupted.
In this mode MySQL will recreate the table, using information from the
@file{.frm} file. This kind of repair cannot be done with @code{myisamchk}.

@node Table maintenance, Maintenance regimen, REPAIR TABLE, Disaster Prevention
@subsection Using @code{myisamchk} for Table Maintenance and Crash Recovery

Starting with MySQL Version 3.23.13, you can check MyISAM
tables with the @code{CHECK TABLE} command. @xref{CHECK TABLE}.  You can
repair tables with the @code{REPAIR TABLE} command. @xref{REPAIR TABLE}.

To check/repair MyISAM tables (@file{.MYI} and @file{.MYD}) you should
use the @code{myisamchk} utility. To check/repair ISAM tables
(@file{.ISM} and @file{.ISD}) you should use the @code{isamchk}
utility. @xref{Table types}.

In the following text we will talk about @code{myisamchk}, but everything
also applies to the old @code{isamchk}.

You can use the @code{myisamchk} utility to get information about your
database tables, check and repair them, or optimise them.  The following
sections describe how to invoke @code{myisamchk} (including a
description of its options), how to set up a table maintenance schedule,
and how to use @code{myisamchk} to perform its various functions.

You can, in most cases, also use the command @code{OPTIMIZE TABLES} to
optimise and repair tables, but this is not as fast or reliable (in case
of real fatal errors) as @code{myisamchk}.  On the other hand,
@code{OPTIMIZE TABLE} is easier to use and you don't have to worry about
flushing tables.
@xref{OPTIMIZE TABLE, , @code{OPTIMIZE TABLE}}.

Even that the repair in @code{myisamchk} is quite secure, it's always a
good idea to make a backup @emph{before} doing a repair (or anything that could
make a lot of changes to a table)

@menu
* myisamchk syntax::            @code{myisamchk} Invocation Syntax
* myisamchk general options::   General Options for @code{myisamchk}
* myisamchk check options::     Check Options for @code{myisamchk}
* myisamchk repair options::    Repair Options for myisamchk
* myisamchk other options::     Other Options for @code{myisamchk}
* myisamchk memory::            @code{myisamchk} Memory Usage
* Crash recovery::              Using @code{myisamchk} for Crash Recovery
* Check::                       How to Check Tables for Errors
* Repair::                      How to Repair Tables
* Optimisation::                Table Optimisation
@end menu


@node myisamchk syntax, myisamchk general options, Table maintenance, Table maintenance
@subsubsection @code{myisamchk} Invocation Syntax

@code{myisamchk} is invoked like this:

@example
shell> myisamchk [options] tbl_name
@end example

The @code{options} specify what you want @code{myisamchk} to do.  They are
described here.  (You can also get a list of options by invoking
@code{myisamchk --help}.)  With no options, @code{myisamchk} simply checks your
table.  To get more information or to tell @code{myisamchk} to take corrective
action, specify options as described here and in the following sections.

@code{tbl_name} is the database table you want to check/repair.  If you run
@code{myisamchk} somewhere other than in the database directory, you must
specify the path to the file, because @code{myisamchk} has no idea where your
database is located.  Actually, @code{myisamchk} doesn't care whether
the files you are working on are located in a database directory; you can
copy the files that correspond to a database table into another location and
perform recovery operations on them there.

You can name several tables on the @code{myisamchk} command-line if you
wish.  You can also specify a name as an index file
name (with the @file{.MYI} suffix), which allows you to specify all
tables in a directory by using the pattern @file{*.MYI}.
For example, if you are in a database directory, you can check all the
tables in the directory like this:

@example
shell> myisamchk *.MYI
@end example

If you are not in the database directory, you can check all the tables there
by specifying the path to the directory:

@example
shell> myisamchk /path/to/database_dir/*.MYI
@end example

You can even check all tables in all databases by specifying a wildcard
with the path to the MySQL data directory:

@example
shell> myisamchk /path/to/datadir/*/*.MYI
@end example

The recommended way to quickly check all tables is:

@example
myisamchk --silent --fast /path/to/datadir/*/*.MYI
isamchk --silent /path/to/datadir/*/*.ISM
@end example

If you want to check all tables and repair all tables that are corrupted,
you can use the following line:

@example
myisamchk --silent --force --fast --update-state -O key_buffer=64M \
          -O sort_buffer=64M -O read_buffer=1M -O write_buffer=1M \
          /path/to/datadir/*/*.MYI
isamchk --silent --force -O key_buffer=64M -O sort_buffer=64M \
        -O read_buffer=1M -O write_buffer=1M /path/to/datadir/*/*.ISM
@end example

The above assumes that you have more than 64 M free.

Note that if you get an error like:

@example
myisamchk: warning: 1 clients is using or hasn't closed the table properly
@end example

This means that you are trying to check a table that has been updated by
the another program (like the @code{mysqld} server) that hasn't yet closed
the file or that has died without closing the file properly.

If you @code{mysqld} is running, you must force a sync/close of all
tables with @code{FLUSH TABLES} and ensure that no one is using the
tables while you are running @code{myisamchk}.  In MySQL Version 3.23
the easiest way to avoid this problem is to use @code{CHECK TABLE}
instead of @code{myisamchk} to check tables.

@node myisamchk general options, myisamchk check options, myisamchk syntax, Table maintenance
@subsubsection General Options for @code{myisamchk}

@cindex options, @code{myisamchk}
@cindex @code{myisamchk}, options

@code{myisamchk} supports the following options.

@table @code
@item -# or --debug=debug_options
Output debug log. The @code{debug_options} string often is
@code{'d:t:o,filename'}.
@item -? or --help
Display a help message and exit.
@item -O var=option, --set-variable var=option
Set the value of a variable.  The possible variables and their default values
for myisamchk can be examined with @code{myisamchk --help}:
@multitable @columnfractions .20 .10
@item @strong{Variable} @tab @strong{Value}
@item key_buffer_size @tab 523264
@item read_buffer_size @tab 262136
@item write_buffer_size @tab 262136
@item sort_buffer_size @tab 2097144
@item sort_key_blocks @tab 16
@item decode_bits @tab 9
@end multitable

@code{sort_buffer_size} is used when the keys are repaired by sorting
keys, which is the normal case when you use @code{--recover}.

@code{key_buffer_size} is used when you are checking the table with
@code{--extended-check} or when the keys are repaired by inserting key
row by row in to the table (like when doing normal inserts). Repairing
through the key buffer is used in the following cases:

@itemize @bullet
@item
If you use @code{--safe-recover}.
@item
If the temporary files needed to sort the keys would be more than twice
as big as when creating the key file directly.  This is often the case
when you have big @code{CHAR}, @code{VARCHAR} or @code{TEXT} keys as the
sort needs to store the whole keys during sorting. If you have lots
of temporary space and you can force @code{myisamchk} to repair by sorting
you can use the @code{--sort-recover} option.
@end itemize

Reparing through the key buffer takes much less disk space than using
sorting, but is also much slower.

If you want a faster repair, set the above variables to about 1/4 of your
available memory.  You can set both variables to big values, as only one
of the above buffers will be used at a time.

@item -s or --silent
Silent mode.  Write output only when errors occur. You can use @code{-s}
twice (@code{-ss}) to make @code{myisamchk} very silent.
@item -v or --verbose
Verbose mode.  Print more information. This can be used with @code{-d} and
@code{-e}. Use @code{-v} multiple times (@code{-vv}, @code{-vvv}) for more
verbosity!
@item -V or --version
Print the @code{myisamchk} version and exit.
@item -w or, --wait
Instead of giving an error if the table is locked, wait until the table
is unlocked before continuing.  Note that if you are running @code{mysqld}
on the table with @code{--skip-locking}, the table can only be locked
by another @code{myisamchk} command.
@end table


@node myisamchk check options, myisamchk repair options, myisamchk general options, Table maintenance
@subsubsection Check Options for @code{myisamchk}

@cindex check options, myisamchk
@cindex tables, checking

@table @code
@item -c or --check
Check table for errors. This is the default operation if you are not
giving @code{myisamchk} any options that override this.

@item -e or --extend-check
Check the table very thoroughly (which is quite slow if you have many
indexes).  This option should only be used in extreme cases.  Normally,
@code{myisamchk} or @code{myisamchk --medium-check} should, in most
cases, be able to find out if there are any errors in the table.

If you are using @code{--extended-check} and have much memory, you should
increase the value of @code{key_buffer_size} a lot!

@item -F or --fast
Check only tables that haven't been closed properly.
@item -C or --check-only-changed
Check only tables that have changed since the last check.
@item -f or --force
Restart @code{myisamchk} with @code{-r} (repair) on the table, if
@code{myisamchk} finds any errors in the table.
@item -i or --information
Print informational statistics about the table that is checked.
@item -m or --medium-check
Faster than extended-check, but only finds 99.99% of all errors.
Should, however, be good enough for most cases.
@item -U or --update-state
Store in the @file{.MYI} file when the table was checked and if the table crashed.  This should be used to get full benefit of the
@code{--check-only-changed} option, but you shouldn't use this
option if the @code{mysqld} server is using the table and you are
running @code{mysqld} with @code{--skip-locking}.
@item -T or --read-only
Don't mark table as checked. This is useful if you use @code{myisamchk}
to check a table that is in use by some other application that doesn't
use locking (like @code{mysqld --skip-locking}).
@end table


@node myisamchk repair options, myisamchk other options, myisamchk check options, Table maintenance
@subsubsection Repair Options for myisamchk

@cindex repair options, myisamchk
@cindex files, repairing

The following options are used if you start @code{myisamchk} with
@code{-r} or @code{-o}:

@table @code
@item -D # or --data-file-length=#
Max length of datafile (when re-creating datafile when it's 'full').
@item -e or --extend-check
Try to recover every possible row from the datafile.
Normally this will also find a lot of garbage rows. Don't use this option
if you are not totally desperate.
@item -f or --force
Overwrite old temporary files (@code{table_name.TMD}) instead of aborting.
@item -k # or keys-used=#
If you are using ISAM, tells the ISAM table handler to update only the
first @code{#} indexes.  If you are using @code{MyISAM}, tells which keys
to use, where each binary bit stands for one key (first key is bit 0).
This can be used to get faster inserts!  Deactivated indexes can be
reactivated by using @code{myisamchk -r}.  keys.
@item -l or --no-symlinks
Do not follow symbolic links. Normally @code{myisamchk} repairs the
table a symlink points at.  This option doesn't exist in MySQL 4.0,
as MySQL 4.0 will not remove symlinks during repair.
@item -r or --recover
Can fix almost anything except unique keys that aren't unique
(which is an extremely unlikely error with ISAM/MyISAM tables).
If you want to recover a table, this is the option to try first. Only if
myisamchk reports that the table can't be recovered by @code{-r}, you
should then try @code{-o}.  (Note that in the unlikely case that @code{-r}
fails, the datafile is still intact.)
If you have lots of memory, you should increase the size of
@code{sort_buffer_size}!
@item  -o or --safe-recover
Uses an old recovery method (reads through all rows in order and updates
all index trees based on the found rows); this is an order of magnitude
slower than @code{-r}, but can handle a couple of very unlikely cases that
@code{-r} cannot handle.  This recovery method also uses much less disk
space than @code{-r}. Normally one should always first repair with
@code{-r}, and only if this fails use @code{-o}.

If you have lots of memory, you should increase the size of
@code{key_buffer_size}!
@item -n or --sort-recover
Force @code{myisamchk} to use sorting to resolve the keys even if the
temporary files should be very big.

@item --character-sets-dir=...
Directory where character sets are stored.
@item --set-character-set=name
Change the character set used by the index
@item -t or --tmpdir=path
Path for storing temporary files. If this is not set, @code{myisamchk} will
use the environment variable @code{TMPDIR} for this.
@item -q or --quick
Faster repair by not modifying the datafile. One can give a second
@code{-q} to force @code{myisamchk} to modify the original datafile in case
of duplicate keys
@item -u or --unpack
Unpack file packed with myisampack.
@end table


@node myisamchk other options, myisamchk memory, myisamchk repair options, Table maintenance
@subsubsection Other Options for @code{myisamchk}

Other actions that @code{myisamchk} can do, besides repair and check tables:

@table @code
@item -a or --analyze
Analyse the distribution of keys. This improves join performance by
enabling the join optimiser to better choose in which order it should
join the tables and which keys it should use:
@code{myisamchk --describe --verbose table_name'} or using @code{SHOW KEYS} in
MySQL.
@item -d or --description
Prints some information about table.
@item -A or --set-auto-increment[=value]
Force @code{AUTO_INCREMENT} to start at this or higher value. If no value is
given, then sets the next @code{AUTO_INCREMENT} value to the highest used value
for the auto key + 1.
@item -S or --sort-index
Sort the index tree blocks in high-low order.
This will optimise seeks and will make table scanning by key faster.
@item -R or --sort-records=#
Sorts records according to an index.  This makes your data much more localised
and may speed up ranged @code{SELECT} and @code{ORDER BY} operations on
this index. (It may be very slow to do a sort the first time!)
To find out a table's index numbers, use @code{SHOW INDEX}, which shows a
table's indexes in the same order that @code{myisamchk} sees them.  Indexes are
numbered beginning with 1.
@end table


@node myisamchk memory, Crash recovery, myisamchk other options, Table maintenance
@subsubsection @code{myisamchk} Memory Usage

@cindex memory usage, myisamchk

Memory allocation is important when you run @code{myisamchk}.
@code{myisamchk} uses no more memory than you specify with the @code{-O}
options.  If you are going to use @code{myisamchk} on very large files,
you should first decide how much memory you want it to use.  The default
is to use only about 3M to fix things.  By using larger values, you can
get @code{myisamchk} to operate faster.  For example, if you have more
than 32M RAM, you could use options such as these (in addition to any
other options you might specify):

@example
shell> myisamchk -O sort=16M -O key=16M -O read=1M -O write=1M ...
@end example

Using @code{-O sort=16M} should probably be enough for most cases.

Be aware that @code{myisamchk} uses temporary files in @code{TMPDIR}. If
@code{TMPDIR} points to a memory filesystem, you may easily get out of
memory errors. If this happens, set @code{TMPDIR} to point at some directory
with more space and restart @code{myisamchk}.

When repairing, @code{myisamchk} will also need a lot of disk space:

@itemize @bullet
@item
Double the size of the record file (the original one and a copy).  This
space is not needed if one does a repair with @code{--quick}, as in this
case only the index file will be re-created.  This space is needed on the
same disk as the original record file!
@item
Space for the new index file that replaces the old one. The old
index file is truncated at start, so one usually ignore this space.
This space is needed on the same disk as the original index file!
@item
When using @code{--recover} or @code{--sort-recover}
(but not when using @code{--safe-recover}), you will need space for a
sort buffer for:
@code{(largest_key + row_pointer_length)*number_of_rows * 2}.
You can check the length of the keys and the row_pointer_length with
@code{myisamchk -dv table}.
This space is allocated on the temporary disk (specified by @code{TMPDIR} or
@code{--tmpdir=#}).
@end itemize

If you have a problem with disk space during repair, you can try to use
@code{--safe-recover} instead of @code{--recover}.


@node Crash recovery, Check, myisamchk memory, Table maintenance
@subsubsection Using @code{myisamchk} for Crash Recovery

@cindex crash, recovery
@cindex recovery, from crash

If you run @code{mysqld} with @code{--skip-locking} (which is the default on
some systems, like Linux), you can't reliably use @code{myisamchk} to
check a table when @code{mysqld} is using the same table.  If you
can be sure that no one is accessing the tables through @code{mysqld}
while you run @code{myisamchk}, you only have to do @code{mysqladmin
flush-tables} before you start checking the tables.  If you can't
guarantee the above, then you must take down @code{mysqld} while you
check the tables.  If you run @code{myisamchk} while @code{mysqld} is updating
the tables, you may get a warning that a table is corrupt even if it
isn't.

If you are not using @code{--skip-locking}, you can use @code{myisamchk}
to check tables at any time.  While you do this, all clients that try
to update the table will wait until @code{myisamchk} is ready before
continuing.

If you use @code{myisamchk} to repair or optimise tables, you
@strong{must} always ensure that the @code{mysqld} server is not using
the table (this also applies if you are using @code{--skip-locking}).
If you don't take down @code{mysqld} you should at least do a
@code{mysqladmin flush-tables} before you run @code{myisamchk}.
Your tables @strong{may be corrupted} if the server and @code{myisamchk}
access the tables simultaneously.

This chapter describes how to check for and deal with data corruption
in MySQL databases.  If your tables get corrupted frequently you should
try to find the reason for this! @xref{Crashing}.

The @code{MyISAM} table section contains reason for why a table could be
corrupted. @xref{MyISAM table problems}.

When performing crash recovery, it is important to understand that each table
@code{tbl_name} in a database corresponds to three files in the database
directory:

@multitable @columnfractions .20 .30
@item @strong{File} @tab @strong{Purpose}
@item @file{tbl_name.frm} @tab Table definition (form) file
@item @file{tbl_name.MYD} @tab Datafile
@item @file{tbl_name.MYI} @tab Index file
@end multitable

Each of these three file types is subject to corruption in various ways, but
problems occur most often in datafiles and index files.

@code{myisamchk} works by creating a copy of the @file{.MYD} (data) file
row by row. It ends the repair stage by removing the old @file{.MYD}
file and renaming the new file to the original file name.  If you use
@code{--quick}, @code{myisamchk} does not create a temporary @file{.MYD}
file, but instead assumes that the @file{.MYD} file is correct and only
generates a new index file without touching the @file{.MYD} file. This
is safe, because @code{myisamchk} automatically detects if the
@file{.MYD} file is corrupt and aborts the repair in this case.  You can
also give two @code{--quick} options to @code{myisamchk}.  In this case,
@code{myisamchk} does not abort on some errors (like duplicate key) but
instead tries to resolve them by modifying the @file{.MYD}
file. Normally the use of two @code{--quick} options is useful only if
you have too little free disk space to perform a normal repair.  In this
case you should at least make a backup before running @code{myisamchk}.


@node Check, Repair, Crash recovery, Table maintenance
@subsubsection How to Check Tables for Errors

@cindex checking, tables for errors
@cindex tables, error checking
@cindex errors, checking tables for

To check a MyISAM table, use the following commands:

@table @code
@item myisamchk tbl_name
This finds 99.99% of all errors. What it can't find is corruption that
involves @strong{only} the datafile (which is very unusual). If you want
to check a table, you should normally run @code{myisamchk} without options or
with either the @code{-s} or @code{--silent} option.

@item myisamchk -m tbl_name
This finds 99.999% of all errors. It checks first all index entries for errors and
then it reads through all rows. It calculates a checksum for all keys in
the rows and verifies that they checksum matches the checksum for the keys
in the index tree.

@item myisamchk -e tbl_name
This does a complete and thorough check of all data (@code{-e} means
``extended check''). It does a check-read of every key for each row to verify
that they indeed point to the correct row.  This may take a long time on a
big table with many keys.  @code{myisamchk} will normally stop after the first
error it finds. If you want to obtain more information, you can add the
@code{--verbose} (@code{-v}) option.  This causes @code{myisamchk} to keep
going, up through a maximum of 20 errors.  In normal usage, a simple
@code{myisamchk} (with no arguments other than the table name) is sufficient.

@item myisamchk -e -i tbl_name
Like the previous command, but the @code{-i} option tells @code{myisamchk} to
print some informational statistics, too.
@end table


@node Repair, Optimisation, Check, Table maintenance
@subsubsection How to Repair Tables

@cindex tables, repairing
@cindex repairing, tables

In the following section we only talk about using @code{myisamchk} on
@code{MyISAM} tables (extensions @file{.MYI} and @file{.MYD}).  If you
are using @code{ISAM} tables (extensions @file{.ISM} and @file{.ISD}),
you should use @code{isamchk} instead.

Starting with MySQL Version 3.23.14, you can repair MyISAM
tables with the @code{REPAIR TABLE} command. @xref{REPAIR TABLE}.

The symptoms of a corrupted table include queries that abort unexpectedly
and observable errors such as these:

@itemize @bullet
@item
@file{tbl_name.frm} is locked against change
@item
Can't find file @file{tbl_name.MYI} (Errcode: ###)
@item
Unexpected end of file
@item
Record file is crashed
@item
Got error ### from table handler

To get more information about the error you can run @code{perror ###}. Here
is the most common errors that indicates a problem with the table:

@example
shell> perror 126 127 132 134 135 136 141 144 145
126 = Index file is crashed / Wrong file format
127 = Record-file is crashed
132 = Old database file
134 = Record was already deleted (or record file crashed)
135 = No more room in record file
136 = No more room in index file
141 = Duplicate unique key or constraint on write or update
144 = Table is crashed and last repair failed
145 = Table was marked as crashed and should be repaired
@end example

Note that error 135, no more room in record file, is not an error that
can be fixed by a simple repair. In this case you have to do:

@example
ALTER TABLE table MAX_ROWS=xxx AVG_ROW_LENGTH=yyy;
@end example

@end itemize

In the other cases, you must repair your tables. @code{myisamchk}
can usually detect and fix most things that go wrong.

The repair process involves up to four stages, described here. Before you
begin, you should @code{cd} to the database directory and check the
permissions of the table files. Make sure they are readable by the Unix user
that @code{mysqld} runs as (and to you, because you need to access the files
you are checking).  If it turns out you need to modify files, they must also
be writable by you.

If you are using MySQL Version 3.23.16 and above, you can (and
should) use the @code{CHECK} and @code{REPAIR} commands to check and repair
@code{MyISAM} tables.  @xref{CHECK TABLE}.  @xref{REPAIR TABLE}.

The manual section about table maintenance includes the options to
@code{isamchk}/@code{myisamchk}.  @xref{Table maintenance}.

The following section is for the cases where the above command fails or
if you want to use the extended features that @code{isamchk}/@code{myisamchk} provides.

If you are going to repair a table from the command-line, you must first
take down the @code{mysqld} server. Note that when you do
@code{mysqladmin shutdown} on a remote server, the @code{mysqld} server
will still be alive for a while after @code{mysqladmin} returns, until
all queries are stopped and all keys have been flushed to disk.

@noindent
@strong{Stage 1: Checking your tables}

Run @code{myisamchk *.MYI} or @code{myisamchk -e *.MYI} if you have
more time.  Use the @code{-s} (silent) option to suppress unnecessary
information.

If the @code{mysqld} server is done you should use the --update option to tell
@code{myisamchk} to mark the table as 'checked'.

You have to repair only those tables for which @code{myisamchk} announces an
error.  For such tables, proceed to Stage 2.

If you get weird errors when checking (such as @code{out of
memory} errors), or if @code{myisamchk} crashes, go to Stage 3.

@noindent
@strong{Stage 2: Easy safe repair}

Note: If you want repairing to go much faster, you should add: @code{-O
sort_buffer=# -O key_buffer=#} (where # is about 1/4 of the available
memory) to all @code{isamchk/myisamchk} commands.

First, try @code{myisamchk -r -q tbl_name} (@code{-r -q} means ``quick
recovery mode''). This will attempt to repair the index file without
touching the datafile.  If the datafile contains everything that it
should and the delete links point at the correct locations within the
datafile, this should work, and the table is fixed. Start repairing the
next table.  Otherwise, use the following procedure:

@enumerate
@item
Make a backup of the datafile before continuing.

@item
Use @code{myisamchk -r tbl_name} (@code{-r} means ``recovery mode''). This will
remove incorrect records and deleted records from the datafile and
reconstruct the index file.

@item
If the preceding step fails, use @code{myisamchk --safe-recover tbl_name}.
Safe recovery mode uses an old recovery method that handles a few cases that
regular recovery mode doesn't (but is slower).
@end enumerate

If you get weird errors when repairing (such as @code{out of
memory} errors), or if @code{myisamchk} crashes, go to Stage 3.

@noindent
@strong{Stage 3: Difficult repair}

You should only reach this stage if the first 16K block in the index file is
destroyed or contains incorrect information, or if the index file is
missing.  In this case, it's necessary to create a new index file. Do so as
follows:

@enumerate
@item
Move the datafile to some safe place.

@item
Use the table description file to create new (empty) data and index files:

@example
shell> mysql db_name
mysql> SET AUTOCOMMIT=1;
mysql> TRUNCATE TABLE table_name;
mysql> quit
@end example

If your SQL version doesn't have @code{TRUNCATE TABLE}, use @code{DELETE FROM
table_name} instead.

@item
Copy the old datafile back onto the newly created datafile.
(Don't just move the old file back onto the new file; you want to retain
a copy in case something goes wrong.)
@end enumerate

Go back to Stage 2.  @code{myisamchk -r -q} should work now.  (This shouldn't
be an endless loop.)

As of @code{MySQL} 4.0.2 you can also use @code{REPAIR ... USE_FRM}
which performs the whole procedure automatically.

@noindent
@strong{Stage 4: Very difficult repair}

You should reach this stage only if the description file has also
crashed. That should never happen, because the description file isn't changed
after the table is created:

@enumerate
@item
Restore the description file from a backup and go back to Stage 3.  You can
also restore the index file and go back to Stage 2.  In the latter case, you
should start with @code{myisamchk -r}.

@item
If you don't have a backup but know exactly how the table was created, create
a copy of the table in another database.  Remove the new datafile, then move
the description and index files from the other database to your crashed
database.  This gives you new description and index files, but leaves
the datafile alone.  Go back to Stage 2 and attempt to reconstruct
the index file.
@end enumerate


@node Optimisation,  , Repair, Table maintenance
@subsubsection Table Optimisation

@cindex tables, optimising
@cindex optimising, tables

To coalesce fragmented records and eliminate wasted space resulting from
deleting or updating records, run @code{myisamchk} in recovery mode:

@example
shell> myisamchk -r tbl_name
@end example

You can optimise a table in the same way using the SQL @code{OPTIMIZE TABLE}
statement.  @code{OPTIMIZE TABLE} does a repair of the table and a key
analysis, and also sorts the index tree to give faster key lookups.
There is also no possibility of unwanted interaction between a utility
and the server, because the server does all the work when you use
@code{OPTIMIZE TABLE}. @xref{OPTIMIZE TABLE}.

@code{myisamchk} also has a number of other options you can use to improve
the performance of a table:

@itemize
@item @code{-S}, @code{--sort-index}
@item @code{-R index_num}, @code{--sort-records=index_num}
@item @code{-a}, @code{--analyze}
@end itemize

For a full description of the option. @xref{myisamchk syntax}.


@node Maintenance regimen, Table-info, Table maintenance, Disaster Prevention
@subsection Setting Up a Table Maintenance Regimen

@cindex maintaining, tables
@cindex tables, maintenance regimen

Starting with MySQL Version 3.23.13, you can check MyISAM
tables with the @code{CHECK TABLE} command. @xref{CHECK TABLE}.  You can
repair tables with the @code{REPAIR TABLE} command. @xref{REPAIR TABLE}.

It is a good idea to perform table checks on a regular basis rather than
waiting for problems to occur.  For maintenance purposes, you can use
@code{myisamchk -s} to check tables.  The @code{-s} option (short for
@code{--silent}) causes @code{myisamchk} to run in silent mode, printing
messages only when errors occur.

@tindex .pid (process ID) file
It's also a good idea to check tables when the server starts up.
For example, whenever the machine has done a reboot in the middle of an
update, you usually need to check all the tables that could have been
affected. (This is an ``expected crashed table''.) You could add a test to
@code{safe_mysqld} that runs @code{myisamchk} to check all tables that have
been modified during the last 24 hours if there is an old @file{.pid}
(process ID) file left after a reboot.  (The @file{.pid} file is created by
@code{mysqld} when it starts up and removed when it terminates normally.  The
presence of a @file{.pid} file at system startup time indicates that
@code{mysqld} terminated abnormally.)

An even better test would be to check any table whose last-modified time
is more recent than that of the @file{.pid} file.

You should also check your tables regularly during normal system
operation.  At MySQL AB, we run a @code{cron} job to check all
our important tables once a week, using a line like this in a @file{crontab}
file:

@example
35 0 * * 0 /path/to/myisamchk --fast --silent /path/to/datadir/*/*.MYI
@end example

This prints out information about crashed tables so we can examine and repair
them when needed.

As we haven't had any unexpectedly crashed tables (tables that become
corrupted for reasons other than hardware trouble)
for a couple of years now (this is really true), once a week is
more than enough for us.

We recommend that to start with, you execute @code{myisamchk -s} each
night on all tables that have been updated during the last 24 hours,
until you come to trust MySQL as much as we do.

@cindex tables, defragment
Normally you don't need to maintain MySQL tables that much.  If
you are changing tables with dynamic size rows (tables with @code{VARCHAR},
@code{BLOB} or @code{TEXT} columns) or have tables with many deleted rows
you may want to from time to time (once a month?) defragment/reclaim space
from the tables.

You can do this by using @code{OPTIMIZE TABLE} on the tables in question or
if you can take the @code{mysqld} server down for a while do:

@example
isamchk -r --silent --sort-index -O sort_buffer_size=16M */*.ISM
myisamchk -r --silent --sort-index  -O sort_buffer_size=16M */*.MYI
@end example


@node Table-info,  , Maintenance regimen, Disaster Prevention
@subsection Getting Information About a Table

@cindex tables, information

To get a description of a table or statistics about it, use the commands shown
here. We explain some of the information in more detail later:

@itemize @bullet
@item myisamchk -d tbl_name
Runs @code{myisamchk} in ``describe mode'' to produce a description of
your table. If you start the MySQL server using the
@code{--skip-locking} option, @code{myisamchk} may report an error for a
table that is updated while it runs.  However, because @code{myisamchk}
doesn't change the table in describe mode, there isn't any risk of
destroying data.

@item myisamchk -d -v tbl_name
To produce more information about what @code{myisamchk} is doing, add @code{-v}
to tell it to run in verbose mode.

@item myisamchk -eis tbl_name
Shows only the most important information from a table. It is slow because it
must read the whole table.

@item myisamchk -eiv tbl_name
This is like @code{-eis}, but tells you what is being done.
@end itemize

@cindex examples, @code{myisamchk} output
@cindex @code{myisamchk}, example output
Example of @code{myisamchk -d} output:
@example
MyISAM file:     company.MYI
Record format:   Fixed length
Data records:    1403698  Deleted blocks:         0
Recordlength:    226

table description:
Key Start Len Index   Type
1   2     8   unique  double
2   15    10  multip. text packed stripped
3   219   8   multip. double
4   63    10  multip. text packed stripped
5   167   2   multip. unsigned short
6   177   4   multip. unsigned long
7   155   4   multip. text
8   138   4   multip. unsigned long
9   177   4   multip. unsigned long
    193   1           text
@end example

Example of @code{myisamchk -d -v} output:
@example
MyISAM file:         company
Record format:       Fixed length
File-version:        1
Creation time:       1999-10-30 12:12:51
Recover time:        1999-10-31 19:13:01
Status:              checked
Data records:           1403698  Deleted blocks:              0
Datafile parts:         1403698  Deleted data:                0
Datafilepointer (bytes):      3  Keyfile pointer (bytes):     3
Max datafile length: 3791650815  Max keyfile length: 4294967294
Recordlength:               226

table description:
Key Start Len Index   Type                  Rec/key     Root Blocksize
1   2     8   unique  double                      1 15845376      1024
2   15    10  multip. text packed stripped        2 25062400      1024
3   219   8   multip. double                     73 40907776      1024
4   63    10  multip. text packed stripped        5 48097280      1024
5   167   2   multip. unsigned short           4840 55200768      1024
6   177   4   multip. unsigned long            1346 65145856      1024
7   155   4   multip. text                     4995 75090944      1024
8   138   4   multip. unsigned long              87 85036032      1024
9   177   4   multip. unsigned long             178 96481280      1024
    193   1           text
@end example

Example of @code{myisamchk -eis} output:
@example
Checking MyISAM file: company
Key:  1:  Keyblocks used:  97%  Packed:    0%  Max levels:  4
Key:  2:  Keyblocks used:  98%  Packed:   50%  Max levels:  4
Key:  3:  Keyblocks used:  97%  Packed:    0%  Max levels:  4
Key:  4:  Keyblocks used:  99%  Packed:   60%  Max levels:  3
Key:  5:  Keyblocks used:  99%  Packed:    0%  Max levels:  3
Key:  6:  Keyblocks used:  99%  Packed:    0%  Max levels:  3
Key:  7:  Keyblocks used:  99%  Packed:    0%  Max levels:  3
Key:  8:  Keyblocks used:  99%  Packed:    0%  Max levels:  3
Key:  9:  Keyblocks used:  98%  Packed:    0%  Max levels:  4
Total:    Keyblocks used:  98%  Packed:   17%

Records:          1403698    M.recordlength:     226
Packed:             0%
Recordspace used:     100%   Empty space:          0%
Blocks/Record:   1.00
Record blocks:    1403698    Delete blocks:        0
Recorddata:     317235748    Deleted data:         0
Lost space:             0    Linkdata:             0

User time 1626.51, System time 232.36
Maximum resident set size 0, Integral resident set size 0
Non physical pagefaults 0, Physical pagefaults 627, Swaps 0
Blocks in 0 out 0, Messages in 0 out 0, Signals 0
Voluntary context switches 639, Involuntary context switches 28966
@end example

Example of @code{myisamchk -eiv} output:
@example
Checking MyISAM file: company
Data records: 1403698   Deleted blocks:       0
- check file-size
- check delete-chain
block_size 1024:
index  1:
index  2:
index  3:
index  4:
index  5:
index  6:
index  7:
index  8:
index  9:
No recordlinks
- check index reference
- check data record references index: 1
Key:  1:  Keyblocks used:  97%  Packed:    0%  Max levels:  4
- check data record references index: 2
Key:  2:  Keyblocks used:  98%  Packed:   50%  Max levels:  4
- check data record references index: 3
Key:  3:  Keyblocks used:  97%  Packed:    0%  Max levels:  4
- check data record references index: 4
Key:  4:  Keyblocks used:  99%  Packed:   60%  Max levels:  3
- check data record references index: 5
Key:  5:  Keyblocks used:  99%  Packed:    0%  Max levels:  3
- check data record references index: 6
Key:  6:  Keyblocks used:  99%  Packed:    0%  Max levels:  3
- check data record references index: 7
Key:  7:  Keyblocks used:  99%  Packed:    0%  Max levels:  3
- check data record references index: 8
Key:  8:  Keyblocks used:  99%  Packed:    0%  Max levels:  3
- check data record references index: 9
Key:  9:  Keyblocks used:  98%  Packed:    0%  Max levels:  4
Total:    Keyblocks used:   9%  Packed:   17%

- check records and index references
[LOTS OF ROW NUMBERS DELETED]

Records:          1403698    M.recordlength:     226   Packed:             0%
Recordspace used:     100%   Empty space:          0%  Blocks/Record:   1.00
Record blocks:    1403698    Delete blocks:        0
Recorddata:     317235748    Deleted data:         0
Lost space:             0    Linkdata:             0

User time 1639.63, System time 251.61
Maximum resident set size 0, Integral resident set size 0
Non physical pagefaults 0, Physical pagefaults 10580, Swaps 0
Blocks in 4 out 0, Messages in 0 out 0, Signals 0
Voluntary context switches 10604, Involuntary context switches 122798
@end example

Here are the sizes of the data and index files for the table used in the
preceding examples:

@example
-rw-rw-r--   1 monty    tcx     317235748 Jan 12 17:30 company.MYD
-rw-rw-r--   1 davida   tcx      96482304 Jan 12 18:35 company.MYM
@end example

Explanations for the types of information @code{myisamchk} produces are
given here.  The ``keyfile'' is the index file.  ``Record'' and ``row''
are synonymous:

@itemize @bullet
@item ISAM file
Name of the ISAM (index) file.

@item Isam-version
Version of ISAM format. Currently always 2.

@item Creation time
When the datafile was created.

@item Recover time
When the index/datafile was last reconstructed.

@item Data records
How many records are in the table.

@item Deleted blocks
How many deleted blocks still have reserved space.
You can optimise your table to minimise this space.
@xref{Optimisation}.

@item Data file: Parts
For dynamic record format, this indicates how many data blocks there are. For
an optimised table without fragmented records, this is the same as @code{Data
records}.

@item Deleted data
How many bytes of non-reclaimed deleted data there are.
You can optimise your table to minimise this space.
@xref{Optimisation}.

@item Data file pointer
The size of the datafile pointer, in bytes. It is usually 2, 3, 4, or 5
bytes. Most tables manage with 2 bytes, but this cannot be controlled
from MySQL yet. For fixed tables, this is a record address. For
dynamic tables, this is a byte address.

@item Keyfile pointer
The size of the index file pointer, in bytes. It is usually 1, 2, or 3
bytes. Most tables manage with 2 bytes, but this is calculated
automatically by MySQL. It is always a block address.

@item Max datafile length
How long the table's datafile (@file{.MYD} file) can become, in bytes.

@item Max keyfile length
How long the table's key file (@file{.MYI} file) can become, in bytes.

@item Recordlength
How much space each record takes, in bytes.

@item Record format
The format used to store table rows.
The examples shown above use @code{Fixed length}.
Other possible values are @code{Compressed} and @code{Packed}.

@item table description
A list of all keys in the table. For each key, some low-level information
is presented:

@itemize @minus
@item Key
This key's number.

@item Start
Where in the record this index part starts.

@item Len
How long this index part is. For packed numbers, this should always be
the full length of the column. For strings, it may be shorter than the full
length of the indexed column, because you can index a prefix of a string
column.

@item Index
@code{unique} or @code{multip.} (multiple). Indicates whether one value
can exist multiple times in this index.

@item Type
What data-type this index part has. This is an ISAM data-type
with the options @code{packed}, @code{stripped} or @code{empty}.

@item Root
Address of the root index block.

@item Blocksize
The size of each index block. By default this is 1024, but the value may be
changed at compile time.

@item Rec/key
This is a statistical value used by the optimiser. It tells how many
records there are per value for this key. A unique key always has a
value of 1. This may be updated after a table is loaded (or greatly
changed) with @code{myisamchk -a}. If this is not updated at all, a default
value of 30 is given.
@end itemize

@item
In the first example above, the 9th key is a multi-part key with two parts.

@item Keyblocks used
What percentage of the keyblocks are used. Because the table used in the
examples had just been reorganised with @code{myisamchk}, the values are very
high (very near the theoretical maximum).

@item Packed
MySQL tries to pack keys with a common suffix. This can only be used
for @code{CHAR}/@code{VARCHAR}/@code{DECIMAL} keys. For long strings like
names, this can significantly reduce the space used. In the third example
above, the 4th key is 10 characters long and a 60% reduction in space is
achieved.

@item Max levels
How deep the B-tree for this key is. Large tables with long keys get high
values.

@item Records
How many rows are in the table.

@item M.recordlength
The average record length. For tables with fixed-length records, this is the
exact record length.

@item Packed
MySQL strips spaces from the end of strings. The @code{Packed}
value indicates the percentage of savings achieved by doing this.

@item Recordspace used
What percentage of the datafile is used.

@item Empty space
What percentage of the datafile is unused.

@item Blocks/Record
Average number of blocks per record (that is, how many links a fragmented
record is composed of). This is always 1.0 for fixed-format tables. This
value should stay as close to 1.0 as possible. If it gets too big, you can
reorganise the table with @code{myisamchk}.
@xref{Optimisation}.

@item Recordblocks
How many blocks (links) are used. For fixed format, this is the same as the number
of records.

@item Deleteblocks
How many blocks (links) are deleted.

@item Recorddata
How many bytes in the datafile are used.

@item Deleted data
How many bytes in the datafile are deleted (unused).

@item Lost space
If a record is updated to a shorter length, some space is lost. This is
the sum of all such losses, in bytes.

@item Linkdata
When the dynamic table format is used, record fragments are linked with
pointers (4 to 7 bytes each). @code{Linkdata} is the sum of the amount of
storage used by all such pointers.
@end itemize

If a table has been compressed with @code{myisampack}, @code{myisamchk
-d} prints additional information about each table column.  See
@ref{myisampack, , @code{myisampack}}, for an example of this
information and a description of what it means.


@node Database Administration, Localisation, Disaster Prevention, MySQL Database Administration
@section Database Administration Language Reference


@menu
* OPTIMIZE TABLE::              @code{OPTIMIZE TABLE} Syntax
* ANALYZE TABLE::               @code{ANALYZE TABLE} Syntax
* FLUSH::                       @code{FLUSH} Syntax
* RESET::                       @code{RESET} Syntax
* KILL::                        @code{KILL} Syntax
* SHOW::                        @code{SHOW} Syntax
@end menu


@node OPTIMIZE TABLE, ANALYZE TABLE, Database Administration, Database Administration
@subsection @code{OPTIMIZE TABLE} Syntax

@findex OPTIMIZE TABLE

@cindex tables, defragmenting
@cindex tables, fragmentation

@example
OPTIMIZE TABLE tbl_name[,tbl_name]...
@end example

@code{OPTIMIZE TABLE} should be used if you have deleted a large part of a
table or if you have made many changes to a table with variable-length rows
(tables that have @code{VARCHAR}, @code{BLOB}, or @code{TEXT} columns).
Deleted records are maintained in a linked list and subsequent @code{INSERT}
operations reuse old record positions. You can use @code{OPTIMIZE TABLE} to
reclaim the unused space and to defragment the datafile.

For the moment, @code{OPTIMIZE TABLE} only works on @code{MyISAM} and
@code{BDB} tables. For @code{BDB} tables, @code{OPTIMIZE TABLE} is
currently mapped to @code{ANALYZE TABLE}.
@xref{ANALYZE TABLE, , @code{ANALYZE TABLE}}.

You can get @code{OPTIMIZE TABLE} to work on other table types by starting
@code{mysqld} with @code{--skip-new} or @code{--safe-mode}, but in this
case @code{OPTIMIZE TABLE} is just mapped to @code{ALTER TABLE}.

@code{OPTIMIZE TABLE} works the following way:
@itemize @bullet
@item
If the table has deleted or split rows, repair the table.
@item
If the index pages are not sorted, sort them.
@item
If the statistics are not up to date (and the repair couldn't be done
by sorting the index), update them.
@end itemize

@code{OPTIMIZE TABLE} for a @code{MyISAM} table is equivalent to running
@code{myisamchk --quick --check-only-changed --sort-index --analyze}
on the table.

Note that the table is locked during the time @code{OPTIMIZE TABLE} is
running!


@node ANALYZE TABLE, FLUSH, OPTIMIZE TABLE, Database Administration
@subsection @code{ANALYZE TABLE} Syntax

@findex ANALYZE TABLE

@example
ANALYZE TABLE tbl_name[,tbl_name...]
@end example

Analyse and store the key distribution for the table.  During the
analysis, the table is locked with a read lock.  This works on
@code{MyISAM} and @code{BDB} tables.

This is equivalent to running @code{myisamchk -a} on the table.

MySQL uses the stored key distribution to decide in which order
tables should be joined when one does a join on something else than a
constant.

The command returns a table with the following columns:

@multitable @columnfractions .15 .40
@item @strong{Column} @tab @strong{Value}
@item Table @tab Table name
@item Op @tab Always ``analyze''
@item Msg_type @tab One of @code{status}, @code{error}, @code{info} or @code{warning}.
@item Msg_text @tab The message.
@end multitable

You can check the stored key distribution with the @code{SHOW INDEX} command.
@xref{SHOW DATABASE INFO}.

If the table hasn't changed since the last @code{ANALYZE TABLE} command,
the table will not be analysed again.


@node FLUSH, RESET, ANALYZE TABLE, Database Administration
@subsection @code{FLUSH} Syntax

@findex FLUSH

@cindex @code{mysqladmin}
@cindex clearing, caches
@cindex caches, clearing

@example
FLUSH flush_option [,flush_option] ...
@end example

You should use the @code{FLUSH} command if you want to clear some of the
internal caches MySQL uses.  To execute @code{FLUSH}, you must have
the @strong{reload} privilege.

@code{flush_option} can be any of the following:

@multitable @columnfractions .25 .75
@item @strong{Option} @tab @strong{Description}

@item @code{HOSTS}  @tab Empties the host cache tables.  You should flush the
host tables if some of your hosts change IP number or if you get the
error message @code{Host ... is blocked}.  When more than
@code{max_connect_errors} errors occur in a row for a given host while
connection to the MySQL server, MySQL assumes
something is wrong and blocks the host from further connection requests.
Flushing the host tables allows the host to attempt to connect
again. @xref{Blocked host}. You can start @code{mysqld} with
@code{-O max_connection_errors=999999999} to avoid this error message.

@item @code{DES_KEY_FILE} @tab Reloads the DES keys from the file that was
specified with the @code{--des-key-file} option at server startup time.

@item @code{LOGS} @tab Closes and reopens all log files.
If you have specified the update log file or a binary log file without
an extension, the extension number of the log file will be incremented
by one relative to the previous file.  If you have used an extension in
the file name, MySQL will close and reopen the update log file.
@xref{Update log}. This is the same thing as sending the @code{SIGHUP}
signal to the @code{mysqld} server.

@item @code{PRIVILEGES} @tab Reloads the privileges from the grant tables in
the @code{mysql} database.

@item @code{QUERY CACHE} @tab Defragment the query cache to better utilise its
memory.  This command will not remove any queries from the cache, unlike
@code{RESET QUERY CACHE}.

@item @code{TABLES} @tab Closes all open tables and force all tables in use to be closed. This also flushes the query cache.

@item @code{[TABLE | TABLES] tbl_name [,tbl_name...]} @tab Flushes only the given tables.

@item @code{TABLES WITH READ LOCK} @tab Closes all open tables and locks all tables for all databases with a read until one executes @code{UNLOCK TABLES}. This is very convenient way to get backups if you have a filesystem, like Veritas,that can take snapshots in time.

@item @code{STATUS} @tab Resets most status variables to zero. This is something one should only use when debugging a query.

@item @code{USER_RESOURCES} @tab Resets all user resources to zero. This will enable blocked users to login again. @xref{User resources}.
@end multitable

You can also access each of the commands shown above with the @code{mysqladmin}
utility, using the @code{flush-hosts}, @code{flush-logs}, @code{reload},
or @code{flush-tables} commands.

Take also a look at the @code{RESET} command used with replication.
@xref{RESET, , @code{RESET}}.

@node RESET, KILL, FLUSH, Database Administration
@subsection @code{RESET} Syntax

@example
RESET reset_option [,reset_option] ...
@end example

The @code{RESET} command is used to clear things. It also acts as an stronger
version of the @code{FLUSH} command.  @xref{FLUSH, , @code{FLUSH}}.

To execute @code{RESET}, you must have the @strong{reload} privilege.

@multitable @columnfractions .25 .75
@item @strong{Option} @tab @strong{Description}

@item @code{MASTER} @tab Deletes all binary logs listed in the index file, resetting the binlog
index file to be empty. In pre-3.23.26 versions, @code{FLUSH MASTER} (Master)

@item @code{SLAVE} @tab Makes the slave forget its replication position in the master
logs. In pre 3.23.26 versions the command was called
@code{FLUSH SLAVE}(Slave)

@item @code{QUERY CACHE} @tab Removes all query results from the query cache.
@end multitable

@node KILL, SHOW, RESET, Database Administration
@subsection @code{KILL} Syntax

@findex KILL

@cindex @code{mysqladmin}

@example
KILL thread_id
@end example

Each connection to @code{mysqld} runs in a separate thread.  You can see
which threads are running with the @code{SHOW PROCESSLIST} command and kill
a thread with the @code{KILL thread_id} command.

If you have the @strong{process} privilege, you can see and kill all threads.
Otherwise, you can see and kill only your own threads.

You can also use the @code{mysqladmin processlist} and @code{mysqladmin kill}
commands to examine and kill threads.

When you do a @code{KILL}, a thread specific @code{kill flag} is set for
the thread.

In most cases it may take some time for the thread to die as the kill
flag is only checked at specific intervals.

@itemize @bullet
@item
In @code{SELECT}, @code{ORDER BY} and @code{GROUP BY} loops, the flag is
checked after reading a block of rows. If the kill flag is set the
statement is aborted
@item
When doing an @code{ALTER TABLE} the kill flag is checked before each block of
rows are read from the original table. If the kill flag was set the command
is aborted and the temporary table is deleted.
@item
When doing an @code{UPDATE TABLE} and @code{DELETE TABLE}, the kill flag
is checked after each block read and after each updated or delete
row. If the kill flag is set the statement is aborted.  Note that if you
are not using transactions, the changes will not be rolled back!
@item
@code{GET_LOCK()} will abort with @code{NULL}.
@item
An @code{INSERT DELAYED} thread will quickly flush all rows it has in
memory and die.
@item
If the thread is in the table lock handler (state: @code{Locked}),
the table lock will be quickly aborted.
@item
If the thread is waiting for free disk space in a @code{write} call, the
write is aborted with an disk full error message.
@end itemize

@node SHOW,  , KILL, Database Administration
@subsection @code{SHOW} Syntax

@c FIX more index hits needed

@findex SHOW DATABASE INFO
@findex SHOW DATABASES
@findex SHOW TABLES
@findex SHOW COLUMNS
@findex SHOW FIELDS
@findex SHOW INDEX
@findex SHOW KEYS
@findex SHOW STATUS
@findex SHOW VARIABLES
@findex SHOW PROCESSLIST
@findex SHOW TABLE STATUS
@findex SHOW GRANTS
@findex SHOW CREATE TABLE
@findex SHOW MASTER STATUS
@findex SHOW MASTER LOGS
@findex SHOW SLAVE STATUS

@example
   SHOW DATABASES [LIKE wild]
or SHOW [OPEN] TABLES [FROM db_name] [LIKE wild]
or SHOW [FULL] COLUMNS FROM tbl_name [FROM db_name] [LIKE wild]
or SHOW INDEX FROM tbl_name [FROM db_name]
or SHOW TABLE STATUS [FROM db_name] [LIKE wild]
or SHOW STATUS [LIKE wild]
or SHOW VARIABLES [LIKE wild]
or SHOW LOGS
or SHOW [FULL] PROCESSLIST
or SHOW GRANTS FOR user
or SHOW CREATE TABLE table_name
or SHOW MASTER STATUS
or SHOW MASTER LOGS
or SHOW SLAVE STATUS
@end example

@code{SHOW} provides information about databases, tables, columns, or
status information about the server. If the @code{LIKE wild} part is
used, the @code{wild} string can be a string that uses the SQL @samp{%}
and @samp{_} wildcard characters.

@menu
* SHOW DATABASE INFO::          Retrieving information about Database, Tables, Columns, and Indexes
* SHOW TABLE STATUS::           @code{SHOW TABLE STATUS}
* SHOW STATUS::                 @code{SHOW STATUS}
* SHOW VARIABLES::              @code{SHOW VARIABLES}
* SHOW LOGS::                   @code{SHOW LOGS}
* SHOW PROCESSLIST::            @code{SHOW PROCESSLIST}
* SHOW GRANTS::                 @code{SHOW GRANTS}
* SHOW CREATE TABLE::           @code{SHOW CREATE TABLE}
@end menu


@node SHOW DATABASE INFO, SHOW TABLE STATUS, SHOW, SHOW
@subsubsection Retrieving information about Database, Tables, Columns, and Indexes

@cindex displaying, information, @code{SHOW}

You can use @code{db_name.tbl_name} as an alternative to the @code{tbl_name
FROM db_name} syntax.  These two statements are equivalent:

@example
mysql> SHOW INDEX FROM mytable FROM mydb;
mysql> SHOW INDEX FROM mydb.mytable;
@end example

@code{SHOW DATABASES} lists the databases on the MySQL server
host.  You can also get this list using the @code{mysqlshow} command.

@code{SHOW TABLES} lists the tables in a given database.  You can also
get this list using the @code{mysqlshow db_name} command.

@strong{Note}: if a user doesn't have any privileges for a table, the table
will not show up in the output from @code{SHOW TABLES} or @code{mysqlshow
db_name}.

@code{SHOW OPEN TABLES} lists the tables that are currently open in
the table cache. @xref{Table cache}.  The @code{Comment} field tells
how many times the table is @code{cached} and @code{in_use}.

@code{SHOW COLUMNS} lists the columns in a given table.  If you specify
the @code{FULL} option, you will also get the privileges you have for
each column.  If the column types are different from what you expect them to
be based on a @code{CREATE TABLE} statement, note that MySQL
sometimes changes column types.  @xref{Silent column changes}.

The @code{DESCRIBE} statement provides information similar to
@code{SHOW COLUMNS}.
@xref{DESCRIBE, , @code{DESCRIBE}}.

@code{SHOW FIELDS} is a synonym for @code{SHOW COLUMNS}, and
@code{SHOW KEYS} is a synonym for @code{SHOW INDEX}.  You can also
list a table's columns or indexes with @code{mysqlshow db_name tbl_name}
or @code{mysqlshow -k db_name tbl_name}.

@code{SHOW INDEX} returns the index information in a format that closely
resembles the @code{SQLStatistics} call in ODBC. The following columns
are returned:

@multitable @columnfractions .15 .55
@item @strong{Column}            @tab @strong{Meaning}
@item @code{Table}               @tab Name of the table.
@item @code{Non_unique}          @tab 0 if the index can't contain duplicates.
@item @code{Key_name}            @tab Name of the index.
@item @code{Seq_in_index}        @tab Column sequence number in index,
                                      starting with 1.
@item @code{Column_name}         @tab Column name.
@item @code{Collation}           @tab How the column is sorted in the index.
                                      In MySQL, this can have values
                                      @samp{A} (Ascending) or @code{NULL} (Not
                                      sorted).
@item @code{Cardinality}         @tab Number of unique values in the index.
                                      This is updated by running
                                      @code{isamchk -a}.
@item @code{Sub_part}            @tab Number of indexed characters if the
                                      column is only partly indexed.
                                      @code{NULL} if the entire key is indexed.
@item @code{Null}		 @tab Contains 'YES' if the column may contain @code{NULL}.
@item @code{Index_type}		 @tab Index method used.
@item @code{Comment}             @tab Various remarks.  For now, it tells
in MySQL < 4.0.2 whether index is @code{FULLTEXT} or not.
@end multitable

Note that as the @code{Cardinality} is counted based on statistics
stored as integers, it's not necessarily accurate for small tables.

The @code{Null} and @code{Index_type} columns were added in MySQL 4.0.2.

@node SHOW TABLE STATUS, SHOW STATUS, SHOW DATABASE INFO, SHOW
@subsubsection @code{SHOW TABLE STATUS}

@cindex displaying, table status
@cindex tables, displaying status
@cindex status, tables

@example
SHOW TABLE STATUS [FROM db_name] [LIKE wild]
@end example

@code{SHOW TABLE STATUS} (new in Version 3.23) works likes @code{SHOW
STATUS}, but provides a lot of information about each table. You can
also get this list using the @code{mysqlshow --status db_name} command.
The following columns are returned:

@multitable @columnfractions .20 .70
@item @strong{Column} @tab @strong{Meaning}
@item @code{Name} @tab Name of the table.
@item @code{Type} @tab Type of table. @xref{Table types}.
@item @code{Row_format} @tab The row storage format (Fixed, Dynamic, or Compressed).
@item @code{Rows} @tab Number of rows.
@item @code{Avg_row_length} @tab Average row length.
@item @code{Data_length} @tab Length of the datafile.
@item @code{Max_data_length} @tab Max length of the datafile.
@item @code{Index_length} @tab Length of the index file.
@item @code{Data_free} @tab Number of allocated but not used bytes.
@item @code{Auto_increment} @tab Next autoincrement value.
@item @code{Create_time} @tab When the table was created.
@item @code{Update_time} @tab When the datafile was last updated.
@item @code{Check_time} @tab When the table was last checked.
@item @code{Create_options} @tab Extra options used with @code{CREATE TABLE}.
@item @code{Comment} @tab The comment used when creating the table (or some information why MySQL couldn't access the table information).
@end multitable

@code{InnoDB} tables will report the free space in the tablespace
in the table comment.


@node SHOW STATUS, SHOW VARIABLES, SHOW TABLE STATUS, SHOW
@subsubsection @code{SHOW STATUS}

@cindex @code{mysqladmin}
@code{SHOW STATUS} provides server status information
(like @code{mysqladmin extended-status}). The output resembles that shown
here, though the format and numbers probably differ:

@example
+--------------------------+------------+
| Variable_name            | Value      |
+--------------------------+------------+
| Aborted_clients          | 0          |
| Aborted_connects         | 0          |
| Bytes_received           | 155372598  |
| Bytes_sent               | 1176560426 |
| Connections              | 30023      |
| Created_tmp_disk_tables  | 0          |
| Created_tmp_tables       | 8340       |
| Created_tmp_files        | 60         |
| Delayed_insert_threads   | 0          |
| Delayed_writes           | 0          |
| Delayed_errors           | 0          |
| Flush_commands           | 1          |
| Handler_delete           | 462604     |
| Handler_read_first       | 105881     |
| Handler_read_key         | 27820558   |
| Handler_read_next        | 390681754  |
| Handler_read_prev        | 6022500    |
| Handler_read_rnd         | 30546748   |
| Handler_read_rnd_next    | 246216530  |
| Handler_update           | 16945404   |
| Handler_write            | 60356676   |
| Key_blocks_used          | 14955      |
| Key_read_requests        | 96854827   |
| Key_reads                | 162040     |
| Key_write_requests       | 7589728    |
| Key_writes               | 3813196    |
| Max_used_connections     | 0          |
| Not_flushed_key_blocks   | 0          |
| Not_flushed_delayed_rows | 0          |
| Open_tables              | 1          |
| Open_files               | 2          |
| Open_streams             | 0          |
| Opened_tables            | 44600      |
| Questions                | 2026873    |
| Select_full_join         | 0          |
| Select_full_range_join   | 0          |
| Select_range             | 99646      |
| Select_range_check       | 0          |
| Select_scan              | 30802      |
| Slave_running            | OFF        |
| Slave_open_temp_tables   | 0          |
| Slow_launch_threads      | 0          |
| Slow_queries             | 0          |
| Sort_merge_passes        | 30         |
| Sort_range               | 500        |
| Sort_rows                | 30296250   |
| Sort_scan                | 4650       |
| Table_locks_immediate    | 1920382    |
| Table_locks_waited       | 0          |
| Threads_cached           | 0          |
| Threads_created          | 30022      |
| Threads_connected        | 1          |
| Threads_running          | 1          |
| Uptime                   | 80380      |
+--------------------------+------------+
@end example

@cindex variables, status
The status variables listed above have the following meaning:

@multitable @columnfractions .30 .70
@item @strong{Variable} @tab @strong{Meaning}
@item @code{Aborted_clients} @tab Number of connections aborted because the client died without closing the connection properly. @xref{Communication errors}.
@item @code{Aborted_connects} @tab Number of tries to connect to the MySQL server that failed. @xref{Communication errors}.
@item @code{Bytes_received} @tab Number of bytes received from all clients.
@item @code{Bytes_sent} @tab Number of bytes sent to all clients.
@item @code{Com_xxx} @tab Number of times each xxx command has been executed.
@item @code{Connections} @tab Number of connection attempts to the MySQL server.
@item @code{Created_tmp_disk_tables} @tab Number of implicit temporary tables on disk created while executing statements.
@item @code{Created_tmp_tables} @tab Number of implicit temporary tables in memory created while executing statements.
@item @code{Created_tmp_files} @tab How many temporary files @code{mysqld} has created.
@item @code{Delayed_insert_threads} @tab Number of delayed insert handler threads in use.
@item @code{Delayed_writes} @tab Number of rows written with @code{INSERT DELAYED}.
@item @code{Delayed_errors} @tab Number of rows written with @code{INSERT DELAYED} for which some error occurred (probably @code{duplicate key}).
@item @code{Flush_commands} @tab Number of executed @code{FLUSH} commands.
@item @code{Handler_commit} @tab Number of internal @code{COMMIT} commands.
@item @code{Handler_delete} @tab Number of times a row was deleted from a table.
@item @code{Handler_read_first} @tab Number of times the first entry was read from an index.
If this is high, it suggests that the server is doing a lot of full index scans, for example,
@code{SELECT col1 FROM foo}, assuming that col1 is indexed.
@item @code{Handler_read_key} @tab Number of requests to read a row based on a key. If this
is high, it is a good indication that your queries and tables are properly indexed.
@item @code{Handler_read_next} @tab Number of requests to read next row in key order. This
will be incremented if you are querying an index column with a range constraint. This also
will be incremented if you are doing an index scan.
@item @code{Handler_read_prev} @tab Number of requests to read previous row in key order.  This is mainly used to optimize @code{ORDER BY ... DESC}.
@item @code{Handler_read_rnd} @tab Number of requests to read a row based on a fixed position.
This will be high if you are doing a lot of queries that require sorting of the result.
@item @code{Handler_read_rnd_next} @tab Number of requests to read the next row in the datafile.
This will be high if you are doing a lot of table scans. Generally this suggests that your tables
are not properly indexed or that your queries are not written to take advantage of the indexes you
have.
@item @code{Handler_rollback} @tab  Number of internal @code{ROLLBACK} commands.
@item @code{Handler_update} @tab Number of requests to update a row in a table.
@item @code{Handler_write} @tab Number of requests to insert a row in a table.
@item @code{Key_blocks_used} @tab The number of used blocks in the key cache.
@item @code{Key_read_requests} @tab The number of requests to read a key block from the cache.
@item @code{Key_reads} @tab The number of physical reads of a key block from disk.
@item @code{Key_write_requests} @tab The number of requests to write a key block to the cache.
@item @code{Key_writes} @tab The number of physical writes of a key block to disk.
@item @code{Max_used_connections} @tab The maximum number of connections in use simultaneously.
@item @code{Not_flushed_key_blocks} @tab Keys blocks in the key cache that has changed but hasn't yet been flushed to disk.
@item @code{Not_flushed_delayed_rows} @tab Number of rows waiting to be written in @code{INSERT DELAY} queues.
@item @code{Open_tables} @tab Number of tables that are open.
@item @code{Open_files} @tab  Number of files that are open.
@item @code{Open_streams} @tab Number of streams that are open (used mainly for logging).
@item @code{Opened_tables} @tab Number of tables that have been opened.
@item @code{Rpl_status} @tab Status of failsafe replication. (Not yet in use).
@item @code{Select_full_join} @tab Number of joins without keys (If this is 0, you should carefully check the index of your tables).
@item @code{Select_full_range_join} @tab Number of joins where we used a range search on reference table.
@item @code{Select_range} @tab Number of joins where we used ranges on the first table. (It's normally not critical even if this is big.)
@item @code{Select_scan} @tab Number of joins where we did a full scann of the first table.
@item @code{Select_range_check} @tab Number of joins without keys where we check for key usage after each row (If this is 0, you should carefully check the index of your tables).
@item @code{Questions} @tab Number of queries sent to the server.
@item @code{Slave_open_temp_tables} @tab Number of temporary tables currently
open by the slave thread
@item @code{Slave_running} @tab Is @code{ON} if this is a slave that is connected to a master.
@item @code{Slow_launch_threads} @tab Number of threads that have taken more than @code{slow_launch_time} to create.
@item @code{Slow_queries} @tab Number of queries that have taken more than @code{long_query_time}. @xref{Slow query log}.
@item @code{Sort_merge_passes} @tab Number of merges passes the sort algoritm have had to do. If this value is large you should consider increasing @code{sort_buffer}.
@item @code{Sort_range} @tab Number of sorts that where done with ranges.
@item @code{Sort_rows}  @tab Number of sorted rows.
@item @code{Sort_scan}  @tab Number of sorts that where done by scanning the table.
@item @code{ssl_xxx}	@tab Variables used by SSL; Not yet implemented.
@item @code{Table_locks_immediate} @tab Number of times a table lock was
acquired immediately. Available after 3.23.33.
@item @code{Table_locks_waited} @tab Number of times a table lock could not
be acquired immediately and a wait was needed. If this is high, and you
have performance problems, you should first optimise your queries, and then
either split your table(s) or use replication. Available after 3.23.33.
@item @code{Threads_cached} @tab Number of threads in the thread cache.
@item @code{Threads_connected} @tab Number of currently open connections.
@item @code{Threads_created} @tab Number of threads created to handle connections.
@item @code{Threads_running} @tab Number of threads that are not sleeping.
@item @code{Uptime} @tab How many seconds the server has been up.
@end multitable

Some comments about the above:

@itemize @bullet
@item
If @code{Opened_tables} is big, then your @code{table_cache}
variable is probably too small.
@item
If @code{Key_reads} is big, then your @code{key_buffer_size} variable is
probably too small.  The cache hit rate can be calculated with
@code{Key_reads}/@code{Key_read_requests}.
@item
If @code{Handler_read_rnd} is big, then you probably have a lot of
queries that require MySQL to scan whole tables or you have
joins that don't use keys properly.
@item
If @code{Threads_created} is big, you may want to increase the
@code{thread_cache_size} variable.  The cache hit rate can be calculated
with @code{Threads_created}/@code{Connections}.
@item
If @code{Created_tmp_disk_tables} is big, you may want to increase the
@code{tmp_table_size} variable to get the temporary tables memory-based
instead of disk based.
@end itemize


@node SHOW VARIABLES, SHOW LOGS, SHOW STATUS, SHOW
@subsubsection @code{SHOW VARIABLES}

@example
SHOW VARIABLES [LIKE wild]
@end example

@code{SHOW VARIABLES} shows the values of some MySQL system
variables.  You can also get this information using the @code{mysqladmin
variables} command.  If the default values are unsuitable, you can set most
of these variables using command-line options when @code{mysqld} starts up.
@xref{Command-line options}.

The output resembles that shown here, though the format and numbers may
differ somewhat:

@example
+------------------------------+---------------------------+
| Variable_name                | Value                     |
+------------------------------+---------------------------+
| ansi_mode                    | OFF                       |
| back_log                     | 50                        |
| basedir                      | /my/monty/                |
| bdb_cache_size               | 16777216                  |
| bdb_log_buffer_size          | 32768                     |
| bdb_home                     | /my/monty/data/           |
| bdb_max_lock                 | 10000                     |
| bdb_logdir                   |                           |
| bdb_shared_data              | OFF                       |
| bdb_tmpdir                   | /tmp/                     |
| binlog_cache_size            | 32768                     |
| concurrent_insert            | ON                        |
| connect_timeout              | 5                         |
| datadir                      | /my/monty/data/           |
| delay_key_write              | ON                        |
| delayed_insert_limit         | 100                       |
| delayed_insert_timeout       | 300                       |
| delayed_queue_size           | 1000                      |
| flush                        | OFF                       |
| flush_time                   | 0                         |
| ft_min_word_len              | 4                         |
| ft_max_word_len              | 254                       |
| ft_max_word_len_for_sort     | 20                        |
| ft_boolean_syntax            | + -><()~*                 |
| have_bdb                     | YES                       |
| have_innodb                  | YES                       |
| have_raid                    | YES                       |
| have_openssl                 | NO                        |
| init_file                    |                           |
| interactive_timeout          | 28800                     |
| join_buffer_size             | 131072                    |
| key_buffer_size              | 16776192                  |
| language                     | /my/monty/share/english/  |
| large_files_support          | ON                        |
| log                          | OFF                       |
| log_update                   | OFF                       |
| log_bin                      | OFF                       |
| log_slave_updates            | OFF                       |
| long_query_time              | 10                        |
| low_priority_updates         | OFF                       |
| lower_case_table_names       | 0                         |
| max_allowed_packet           | 1048576                   |
| max_binlog_cache_size        | 4294967295                |
| max_connections              | 100                       |
| max_connect_errors           | 10                        |
| max_delayed_threads          | 20                        |
| max_heap_table_size          | 16777216                  |
| max_join_size                | 4294967295                |
| max_sort_length              | 1024                      |
| max_tmp_tables               | 32                        |
| max_write_lock_count         | 4294967295                |
| myisam_bulk_insert_tree_size | 8388608                   |
| myisam_recover_options       | DEFAULT                   |
| myisam_sort_buffer_size      | 8388608                   |
| net_buffer_length            | 16384                     |
| net_read_timeout             | 30                        |
| net_retry_count              | 10                        |
| net_write_timeout            | 60                        |
| open_files_limit             | 0                         |
| pid_file                     | /my/monty/data/donna.pid  |
| port                         | 3306                      |
| protocol_version             | 10                        |
| record_buffer                | 131072                    |
| query_buffer_size            | 0                         |
| query_cache_limit            | 1048576                   |
| query_cache_size             | 16768060                  |
| query_cache_startup_type     | 1                         |
| safe_show_database           | OFF                       |
| server_id                    | 0                         |
| skip_locking                 | ON                        |
| skip_networking              | OFF                       |
| skip_show_database           | OFF                       |
| slow_launch_time             | 2                         |
| socket                       | /tmp/mysql.sock           |
| sort_buffer                  | 2097116                   |
| table_cache                  | 64                        |
| table_type                   | MYISAM                    |
| thread_cache_size            | 4                         |
| thread_stack                 | 65536                     |
| tmp_table_size               | 1048576                   |
| tmpdir                       | /tmp/                     |
| version                      | 3.23.29a-gamma-debug      |
| wait_timeout                 | 28800                     |
+------------------------------+---------------------------+
@end example

Each option is described here. Values for buffer sizes, lengths, and stack
sizes are given in bytes.  You can specify values with a suffix of @samp{K}
or @samp{M} to indicate kilobytes or megabytes. For example, @code{16M}
indicates 16 megabytes.  The case of suffix letters does not matter;
@code{16M} and @code{16m} are equivalent:

@c FIX 2002-04-29 arjen / paul
@c FIX Below should be a @multitable just like SHOW STATUS.
@c FIX Has to wait till after O'Reilly printed edition is out the door.

@cindex variables, values
@itemize
@item @code{ansi_mode}.
Is @code{ON} if @code{mysqld} was started with @code{--ansi}.
@xref{ANSI mode}.

@item @code{back_log}
The number of outstanding connection requests MySQL can have. This
comes into play when the main MySQL thread gets @strong{very}
many connection requests in a very short time. It then takes some time
(although very little) for the main thread to check the connection and start
a new thread. The @code{back_log} value indicates how many requests can be
stacked during this short time before MySQL momentarily stops
answering new requests. You need to increase this only if you expect a large
number of connections in a short period of time.

In other words, this value is the size of the listen queue for incoming
TCP/IP connections.  Your operating system has its own limit on the size
of this queue.  The manual page for the Unix @code{listen(2)} system
call should have more details.  Check your OS documentation for the
maximum value for this variable.  Attempting to set @code{back_log}
higher than your operating system limit will be ineffective.

@item @code{basedir}
The value of the @code{--basedir} option.

@item @code{bdb_cache_size}
The buffer that is allocated to cache index and rows for @code{BDB}
tables.  If you don't use @code{BDB} tables, you should start
@code{mysqld} with @code{--skip-bdb} to not waste memory for this
cache.

@item @code{bdb_log_buffer_size}
The buffer that is allocated to cache index and rows for @code{BDB}
tables.  If you don't use @code{BDB} tables, you should set this to 0 or
start @code{mysqld} with @code{--skip-bdb} to not waste memory for this
cache.

@item @code{bdb_home}
The value of the @code{--bdb-home} option.

@item @code{bdb_max_lock}
The maximum number of locks (1000 by default) you can have active on a
BDB table. You should increase this if you get errors of type @code{bdb:
Lock table is out of available locks} or @code{Got error 12 from ...}
when you have do long transactions or when @code{mysqld} has to examine
a lot of rows to calculate the query.

@item @code{bdb_logdir}
The value of the @code{--bdb-logdir} option.

@item @code{bdb_shared_data}
Is @code{ON} if you are using @code{--bdb-shared-data}.

@item @code{bdb_tmpdir}
The value of the @code{--bdb-tmpdir} option.

@item @code{binlog_cache_size}.  The size of the cache to hold the SQL
statements for the binary log during a transaction.  If you often use
big, multi-statement transactions you can increase this to get more
performance. @xref{COMMIT}.

@item @code{character_set}
The default character set.

@item @code{character_sets}
The supported character sets.

@item @code{concurrent_inserts}
If @code{ON} (the default), MySQL will allow you to use @code{INSERT} on
@code{MyISAM} tables at the same time as you run @code{SELECT} queries
on them.  You can turn this option off by starting @code{mysqld} with
@code{--safe} or @code{--skip-new}.

@cindex timeout
@item @code{connect_timeout}
The number of seconds the @code{mysqld} server is waiting for a connect
packet before responding with @code{Bad handshake}.

@item @code{datadir}
The value of the @code{--datadir} option.

@item @code{delay_key_write}
If enabled (is on by default), MySQL will honor the
@code{DELAY_KEY_WRITE} option for @code{CREATE TABLE}.  This means that the
key buffer for tables with this option will not get flushed on every
index update, but only when a table is closed.  This will speed up
writes on keys a lot, but you should add automatic checking of all tables
with @code{myisamchk --fast --force} if you use this.  Note that if you
start @code{mysqld} with the @code{--delay-key-write-for-all-tables}
option this means that all tables will be treated as if they were
created with the @code{delay_key_write} option.  You can clear this flag
by starting @code{mysqld} with @code{--skip-new} or @code{--safe-mode}.

@item @code{delayed_insert_limit}
After inserting @code{delayed_insert_limit} rows, the @code{INSERT
DELAYED} handler will check if there are any @code{SELECT} statements
pending. If so, it allows these to execute before continuing.

@item @code{delayed_insert_timeout}
How long a @code{INSERT DELAYED} thread should wait for @code{INSERT}
statements before terminating.

@item @code{delayed_queue_size}
What size queue (in rows) should be allocated for handling @code{INSERT
DELAYED}.  If the queue becomes full, any client that does @code{INSERT
DELAYED} will wait until there is room in the queue again.

@item @code{flush}
This is @code{ON} if you have started MySQL with the @code{--flush}
option.

@item @code{flush_time}
If this is set to a non-zero value, then every @code{flush_time} seconds all
tables will be closed (to free up resources and sync things to disk). We
only recommend this option on Windows 9x/Me, or on systems where you have
very little resources.

@item @code{ft_min_word_len}
The minimum length of the word to be included in a @code{FULLTEXT} index.
@strong{Note: @code{FULLTEXT} indexes must be rebuilt after changing
this variable.} (This option is new for MySQL 4.0.)

@item @code{ft_max_word_len}
The maximum length of the word to be included in a @code{FULLTEXT} index.
@strong{Note: @code{FULLTEXT} indexes must be rebuilt after changing
this variable.} (This option is new for MySQL 4.0.)

@item @code{ft_max_word_len_sort}
The maximum length of the word in a @code{FULLTEXT} index
to be used in fast index recreation method in @code{REPAIR},
@code{CREATE INDEX},  or @code{ALTER TABLE}.  Longer words are inserted the
slow way.  The rule of the thumb is as follows: with
@code{ft_max_word_len_sort} increasing, @strong{MySQL} will create bigger
temporary files (thus slowing the process down, due to disk I/O), and will put
fewer keys in one sort block (again, decreasing the efficiency).  When
@code{ft_max_word_len_sort} is too small, instead, @strong{MySQL} will insert a
lot of words into index the slow way, but short words will be inserted very
quickly.

@item @code{ft_boolean_syntax}
List of operators supported by @code{MATCH ... AGAINST(... IN BOOLEAN MODE)}.
@xref{Fulltext Search}.

@item @code{have_innodb}
@code{YES} if @code{mysqld} supports InnoDB tables. @code{DISABLED}
if @code{--skip-innodb} is used.
@item @code{have_bdb}
@code{YES} if @code{mysqld} supports Berkeley DB tables. @code{DISABLED}
if @code{--skip-bdb} is used.
@item @code{have_raid}
@code{YES} if @code{mysqld} supports the @code{RAID} option.
@item @code{have_openssl}
@code{YES} if @code{mysqld} supports SSL (encryption) on the client/server
protocol.

@item @code{init_file}
The name of the file specified with the @code{--init-file} option when
you start the server.  This is a file of SQL statements you want the
server to execute when it starts.

@item @code{interactive_timeout}
The number of seconds the server waits for activity on an interactive
connection before closing it.  An interactive client is defined as a
client that uses the @code{CLIENT_INTERACTIVE} option to
@code{mysql_real_connect()}.  See also @code{wait_timeout}.

@item @code{join_buffer_size}
The size of the buffer that is used for full joins (joins that do not
use indexes).  The buffer is allocated one time for each full join
between two tables. Increase this value to get a faster full join when
adding indexes is not possible. (Normally the best way to get fast joins
is to add indexes.)

@c Make texi2html support index @anchor{Index cache size}. Then change
@c some xrefs to point here
@cindex indexes, block size
@item @code{key_buffer_size}
Index blocks are buffered and are shared by all threads.
@code{key_buffer_size} is the size of the buffer used for index blocks.

Increase this to get better index handling (for all reads and multiple
writes) to as much as you can afford; 64M on a 256M machine that mainly
runs MySQL is quite common.  If you, however, make this too big
(for instance more than 50% of your total memory) your system may start
to page and become extremely slow. Remember that because MySQL does not
cache data reads, you will have to leave some room for the OS
filesystem cache.

You can check the performance of the key buffer by doing @code{show
status} and examine the variables @code{Key_read_requests},
@code{Key_reads}, @code{Key_write_requests}, and @code{Key_writes}.  The
@code{Key_reads/Key_read_request} ratio should normally be < 0.01.
The @code{Key_write/Key_write_requests} is usually near 1 if you are
using mostly updates/deletes but may be much smaller if you tend to
do updates that affect many at the same time or if you are
using @code{delay_key_write}. @xref{SHOW, , @code{SHOW}}.

To get even more speed when writing many rows at the same time, use
@code{LOCK TABLES}.  @xref{LOCK TABLES, , @code{LOCK TABLES}}.

@item @code{language}
The language used for error messages.

@item @code{large_file_support}
If @code{mysqld} was compiled with options for big file support.

@item @code{locked_in_memory}
If @code{mysqld} was locked in memory with @code{--memlock}

@item @code{log}
If logging of all queries is enabled.

@item @code{log_update}
If the update log is enabled.

@item @code{log_bin}
If the binary log is enabled.

@item @code{log_slave_updates}
If the updates from the slave should be logged.

@item @code{long_query_time}
If a query takes longer than this (in seconds), the @code{Slow_queries} counter
will be incremented. If you are using @code{--log-slow-queries}, the query
will be logged to the slow query logfile. @xref{Slow query log}.

@item @code{lower_case_table_names}
If set to 1 table names are stored in lowercase on disk and table
names will be case-insensitive.
@xref{Name case sensitivity}.

@item @code{max_allowed_packet}
The maximum size of one packet. The message buffer is initialised to
@code{net_buffer_length} bytes, but can grow up to @code{max_allowed_packet}
bytes when needed.  This value by default is small, to catch big (possibly
wrong) packets.  You must increase this value if you are using big
@code{BLOB} columns. It should be as big as the biggest @code{BLOB} you want
to use.  The protocol limits for @code{max_allowed_packet} is 16M in MySQL
3.23 and 2G in MySQL 4.0.

@item @code{max_binlog_cache_size}
If a multi-statement transaction requires more than this amount of memory,
one will get the error "Multi-statement transaction required more than
'max_binlog_cache_size' bytes of storage".

@item @code{max_binlog_size}
Available after 3.23.33. If a write to the binary (replication) log exceeds
the given value, rotate the logs. You cannot set it to less than 1024 bytes,
or more than 1 GB. Default is 1 GB.

@item @code{max_connections}
The number of simultaneous clients allowed. Increasing this value increases
the number of file descriptors that @code{mysqld} requires.  See below for
comments on file descriptor limits. @xref{Too many connections}.

@item @code{max_connect_errors}
If there is more than this number of interrupted connections from a host
this host will be blocked from further connections.  You can unblock a host
with the command @code{FLUSH HOSTS}.

@item @code{max_delayed_threads}
Don't start more than this number of threads to handle @code{INSERT DELAYED}
statements.  If you try to insert data into a new table after all @code{INSERT
DELAYED} threads are in use, the row will be inserted as if the
@code{DELAYED} attribute wasn't specified.

@item @code{max_heap_table_size}
Don't allow creation of heap tables bigger than this.

@item @code{max_join_size}
Joins that are probably going to read more than @code{max_join_size}
records return an error. Set this value if your users tend to perform joins
that lack a @code{WHERE} clause, that take a long time, and that return
millions of rows.

@item @code{max_sort_length}
The number of bytes to use when sorting @code{BLOB} or @code{TEXT}
values (only the first @code{max_sort_length} bytes of each value
are used; the rest are ignored).

@item @code{max_user_connections}
The maximum number of active connections for a single user (0 = no limit).

@item @code{max_tmp_tables}
(This option doesn't yet do anything.)
Maximum number of temporary tables a client can keep open at the same time.

@item @code{max_write_lock_count}
After this many write locks, allow some read locks to run in between.

@item @code{myisam_bulk_insert_tree_size}
MySQL uses special tree-like cache to make bulk inserts (that is,
@code{INSERT ... SELECT}, @code{INSERT ... VALUES (...), (...), ...}, and
@code{LOAD DATA INFILE}) faster. This variable limits
the size of the cache tree in bytes per thread. Setting it to 0
will disable this optimization.
@strong{Note}: this cache is only used when adding data to non-empty table.
Default value is 8 MB.

@item @code{myisam_recover_options}
The value of the @code{--myisam-recover} option.

@item @code{myisam_sort_buffer_size}
The buffer that is allocated when sorting the index when doing a
@code{REPAIR} or when creating indexes with @code{CREATE INDEX} or
@code{ALTER TABLE}.

@item @code{myisam_max_extra_sort_file_size}.
If the temporary file used for fast index creation would be bigger than
using the key cache by the amount specified here, then prefer the key
cache method.  This is mainly used to force long character keys in large
tables to use the slower key cache method to create the index.
@strong{Note} that this parameter is given in megabytes!

@item @code{myisam_max_sort_file_size}
The maximum size of the temporary file MySQL is allowed to use
while recreating the index (during @code{REPAIR}, @code{ALTER TABLE}
or @code{LOAD DATA INFILE}.  If the file-size would be bigger than this,
the index will be created through the key cache (which is slower).
@strong{Note} that this parameter is given in megabytes!

@item @code{net_buffer_length}
The communication buffer is reset to this size between queries. This
should not normally be changed, but if you have very little memory, you
can set it to the expected size of a query.  (That is, the expected length of
SQL statements sent by clients.  If statements exceed this length, the buffer
is automatically enlarged, up to @code{max_allowed_packet} bytes.)

@item @code{net_read_timeout}
Number of seconds to wait for more data from a connection before aborting
the read.  Note that when we don't expect data from a connection, the timeout
is defined by @code{write_timeout}. See also @code{slave_net_timeout}.

@item @code{net_retry_count}
If a read on a communication port is interrupted, retry this many times
before giving up.  This value should be quite high on @code{FreeBSD} as
internal interrupts are sent to all threads.

@item @code{net_write_timeout}
Number of seconds to wait for a block to be written to a connection before
aborting the write.

@item @code{open_files_limit}
If this is not 0, then @code{mysqld} will use this value to reserve file
descriptors to use with @code{setrlimit()}.  If this value is 0 then
@code{mysqld} will reserve @code{max_connections*5} or
@code{max_connections + table_cache*2} (whichever is larger) number of
files.  You should try increasing this if @code{mysqld} gives you the
error 'Too many open files'.

@item @code{pid_file}
The value of the @code{--pid-file} option.

@item @code{port}
The value of the @code{--port} option.

@item @code{protocol_version}
The protocol version used by the MySQL server.

@item @code{record_buffer}
Each thread that does a sequential scan allocates a buffer of this
size for each table it scans. If you do many sequential scans, you may
want to increase this value.

@item @code{record_rnd_buffer}
When reading rows in sorted order after a sort, the rows are read through this
buffer to avoid a disk seeks.  If not set, then it's set to the value of
@code{record_buffer}.

@item @code{query_buffer_size}
The initial allocation of the query buffer. If most of your queries are
long (like when inserting blobs), you should increase this!

@item @code{query_cache_limit}
Don't cache results that are bigger than this. (Default 1M).

@item @code{query_cache_size}
The memory allocated to store results from old queries.
If this is 0, the query cache is disabled (default).

@item @code{query_cache_startup_type}
This may be set (only numeric) to
@multitable @columnfractions .09 .14 .72
@item @strong{Value} @tab @strong{Alias} @tab @strong{Comment}
@item 0 @tab OFF @tab Don't cache or retrieve results.
@item 1 @tab ON @tab Cache all results except @code{SELECT SQL_NO_CACHE ...} queries.
@item 2 @tab DEMAND @tab Cache only @code{SELECT SQL_CACHE ...} queries.
@end multitable

@item @code{safe_show_database}
Don't show databases for which the user doesn't have any database or
table privileges. This can improve security if you're concerned about
people being able to see what databases other users have. See also
@code{skip_show_database}.

@item @code{server_id}
The value of the @code{--server-id} option.

@item @code{skip_locking}
Is OFF if @code{mysqld} uses external locking.

@item @code{skip_networking}
Is ON if we only allow local (socket) connections.

@item @code{skip_show_database}
This prevents people from doing @code{SHOW DATABASES} if they don't have
the @strong{process} privilege. This can improve security if you're
concerned about people being able to see what databases other users
have. See also @code{safe_show_database}.

@item @code{slave_net_timeout}
Number of seconds to wait for more data from a master/slave connection
before aborting the read.

@item @code{slow_launch_time}
If creating the thread takes longer than this value (in seconds), the
@code{Slow_launch_threads} counter will be incremented.

@item @code{socket}
The Unix socket used by the server.

@item @code{sort_buffer}
Each thread that needs to do a sort allocates a buffer of this
size. Increase this value for faster @code{ORDER BY} or @code{GROUP BY}
operations.
@xref{Temporary files}.

@item @code{table_cache}
The number of open tables for all threads. Increasing this value
increases the number of file descriptors that @code{mysqld} requires.
You can check if you need to increase the table cache by checking the
@code{Opened_tables} variable. @xref{SHOW, , @code{SHOW}}.  If this variable
is big and you don't do @code{FLUSH TABLES} a lot (which just forces all
tables to be closed and reopenend), then you should increase the value of this
variable.

For more information about the table cache, see @ref{Table cache}.

@item @code{table_type}
The default table type.

@item @code{thread_cache_size}
How many threads we should keep in a cache for reuse.  When a
client disconnects, the client's threads are put in the cache if there
aren't more than @code{thread_cache_size} threads from before.  All new
threads are first taken from the cache, and only when the cache is empty
is a new thread created.  This variable can be increased to improve
performance if you have a lot of new connections. (Normally this doesn't
give a notable performance improvement if you have a good
thread implementation.)  By examing the difference between
the @code{Connections} and @code{Threads_created} you can see how efficient
the current thread cache is for you.

@item @code{thread_concurrency}
On Solaris, @code{mysqld} will call @code{thr_setconcurrency()} with
this value.  @code{thr_setconcurrency()} permits the application to give
the threads system a hint for the desired number of threads that should
be run at the same time.

@item @code{thread_stack}
The stack size for each thread.  Many of the limits detected by the
@code{crash-me} test are dependent on this value. The default is
large enough for normal operation.  @xref{MySQL Benchmarks}.

@item @code{timezone}
The timezone for the server.

@item @code{tmp_table_size}
If an in-memory temporary table exceeds this size, MySQL
will automatically convert it to an on-disk @code{MyISAM} table.
Increase the value of @code{tmp_table_size} if you do many advanced
@code{GROUP BY} queries and you have lots of memory.

@item @code{tmpdir}
The directory used for temporary files and temporary tables.

@item @code{version}
The version number for the server.

@item @code{wait_timeout}
The number of seconds the server waits for activity on a connection before
closing it. See also @code{interactive_timeout}.
@end itemize

The manual section that describes tuning MySQL contains some
information of how to tune the above variables. @xref{Server parameters}.


@node SHOW LOGS, SHOW PROCESSLIST, SHOW VARIABLES, SHOW
@subsubsection @code{SHOW LOGS}

@code{SHOW LOGS} shows you status information about existing log
files.  It currently only displays information about Berkeley DB log
files.

@itemize @bullet
@item @code{File} shows the full path to the log file
@item @code{Type} shows the type of the log file (@code{BDB} for Berkeley
DB log files)
@item @code{Status} shows the status of the log file (@code{FREE} if the
file can be removed, or @code{IN USE} if the file is needed by the transaction
subsystem)
@end itemize


@node SHOW PROCESSLIST, SHOW GRANTS, SHOW LOGS, SHOW
@subsubsection @code{SHOW PROCESSLIST}

@findex threads
@findex PROCESSLIST

@cindex threads, display
@cindex processes, display

@code{SHOW [FULL] PROCESSLIST} shows you which threads are running.  You can
also get this information using the @code{mysqladmin processlist}
command.  If you have the @strong{process} privilege, you can see all
threads.  Otherwise, you can see only your own threads.  @xref{KILL, ,
@code{KILL}}.  If you don't use the @code{FULL} option, then only
the first 100 characters of each query will be shown.

This command is very useful if you get the 'too many connections' error
message and want to find out what's going on. MySQL reserves
one extra connection for a client with the @strong{process} privilege
to ensure that you should always be able to login and check the system
(assuming you are not giving this privilege to all your users).

Some states commonly seen in @code{mysqladmin processlist}

@itemize @bullet
@item @code{Checking table}
The thread is performing [automatic] checking of the table.
@item @code{Closing tables}
Means that the thread is flushing the changed table data to disk and
closing the used tables. This should be a fast operation.  If not, then
you should check that you don't have a full disk or that the disk is not
in very heavy use.
@item @code{Connect Out}
Slave connecting to master.
@item @code{Copying to tmp table on disk}
The temporary result set was larger than @code{tmp_table_size} and the
thread is now changing the in memory-based temporary table to a disk
based one to save memory.
@item @code{Creating tmp table}
The thread is creating a temporary table to hold a part of the result for
the query.
@item @code{deleting from main table}
When executing the first part of a multi-table delete and we are only
deleting from the first table.
@item @code{deleting from reference tables}
When executing the second part of a multi-table delete and we are deleting
the matched rows from the other tables.
@item @code{Flushing tables}
The thread is executing @code{FLUSH TABLES} and is waiting for all
threads to close their tables.
@item @code{Killed}
Someone has sent a kill to the thread and it should abort next time it
checks the kill flag.  The flag is checked in each major loop in MySQL,
but in some cases it may still take a short time for the thread to die.
If the thread is locked by some other thread, the kill will take affect
as soon as the other thread releases it's lock.
@item @code{Sending data}
The thread is processing rows for a @code{SELECT} statement and is
also sending data to the client.
@item @code{Sorting for group}
The thread is doing a sort to satsify a @code{GROUP BY}.
@item @code{Sorting for order}
The thread is doing a sort to satsify a @code{ORDER BY}.
@item @code{Opening tables}
This simply means that the thread is trying to open a table. This is
should be very fast procedure, unless something prevents opening. For
example an @code{ALTER TABLE} or a @code{LOCK TABLE} can prevent opening
a table until the command is finished.
@item @code{Removing duplicates}
The query was using @code{SELECT DISTINCT} in such a way that MySQL
couldn't optimize that distinct away at an early stage. Because of this
MySQL has to do an extra stage to remove all duplicated rows before
sending the result to the client.
@item @code{Reopen table}
The thread got a lock for the table, but noticed after getting the lock
that the underlying table structure changed. It has freed the lock,
closed the table and is now trying to reopen it.
@item @code{Repair by sorting}
The repair code is using sorting to create indexes.
@item @code{Repair with keycache}
The repair code is using creating keys one by one through the key cache.
This is much slower than @code{Repair by sorting}.
@item @code{Searching rows for update}
The thread is doing a first phase to find all matching rows before
updating them.  This has to be done if the @code{UPDATE} is changing
the index that is used to find the involved rows.
@item @code{Sleeping}
The thread is wating for the client to send a new command to it.
@item @code{System lock}
The thread is waiting for getting to get a external system lock for the
table. If you are not using multiple mysqld servers that are accessing
the same tables, you can disable system locks with the
@code{--skip-locking} option.
@item @code{Upgrading lock}
The @code{INSERT DELAYED} handler is trying to get a lock for the table
to insert rows.
@item @code{Updating}
The thread is searching for rows to update and updating them.
@item @code{User Lock}
The thread is waiting on a @code{GET_LOCK()}.
@item @code{Waiting for tables}
The thread got a notification that the underlying structure for a table
has changed and it needs to reopen the table to get the new structure.
To be able to reopen the table it must however wait until all other
threads have closed the table in question.

This notification happens if another thread has used @code{FLUSH TABLES}
or one of the following commands on the table in question: @code{FLUSH
TABLES table_name}, @code{ALTER TABLE}, @code{RENAME TABLE},
@code{REPAIR TABLE}, @code{ANALYZE TABLE} or @code{OPTIMIZE TABLE}.
@item @code{waiting for handler insert}
The @code{INSERT DELAYED} handler has processed all inserts and are
waiting to get new ones.
@end itemize

Most states are very quick operations. If threads last in any of these
states for many seconds, there may be a problem around that needs to be
investigated.

There are some other states that are not mentioned previously, but most of
these are only useful to find bugs in @code{mysqld}.

@node SHOW GRANTS, SHOW CREATE TABLE, SHOW PROCESSLIST, SHOW
@subsubsection @code{SHOW GRANTS}

@cindex privileges, display

@code{SHOW GRANTS FOR user} lists the grant commands that must be issued to
duplicate the grants for a user.

@example
mysql> SHOW GRANTS FOR root@@localhost;
+---------------------------------------------------------------------+
| Grants for root@@localhost                                           |
+---------------------------------------------------------------------+
| GRANT ALL PRIVILEGES ON *.* TO 'root'@@'localhost' WITH GRANT OPTION |
+---------------------------------------------------------------------+
@end example


@node SHOW CREATE TABLE,  , SHOW GRANTS, SHOW
@subsubsection @code{SHOW CREATE TABLE}

Shows a @code{CREATE TABLE} statement that will create the given table:

@example
mysql> SHOW CREATE TABLE t\G
*************************** 1. row ***************************
       Table: t
Create Table: CREATE TABLE t (
  id int(11) default NULL auto_increment,
  s char(60) default NULL,
  PRIMARY KEY (id)
) TYPE=MyISAM

@end example

@code{SHOW CREATE TABLE} will quote table and column names according to
@code{SQL_QUOTE_SHOW_CREATE} option.
@ref{SET OPTION, , @code{SET OPTION SQL_QUOTE_SHOW_CREATE}}.


@node Localisation, Server-Side Scripts, Database Administration, MySQL Database Administration
@section MySQL Localisation and International Usage

@menu
* Character sets::              The Character Set Used for Data and Sorting
* Languages::                   Non-English Error Messages
* Adding character set::        Adding a New Character Set
* Character arrays::            The Character Definition Arrays
* String collating::            String Collating Support
* Multi-byte characters::       Multi-byte Character Support
* Problems with character sets::  Problems With Character Sets
@end menu


@node Character sets, Languages, Localisation, Localisation
@subsection The Character Set Used for Data and Sorting

@cindex character sets
@cindex data, character sets
@cindex sorting, character sets

By default, MySQL uses the ISO-8859-1 (Latin1) character set
with sorting according to Swedish/Finnish. This is the character set suitable
in the USA and western Europe.

All standard MySQL binaries are compiled with
@code{--with-extra-charsets=complex}.  This will add code to all
standard programs to be able to handle @code{latin1} and all multi-byte
character sets within the binary. Other character sets will be
loaded from a character-set definition file when needed.

The character set determines what characters are allowed in names and how
things are sorted by the @code{ORDER BY} and @code{GROUP BY} clauses of
the @code{SELECT} statement.

You can change the character set with the @code{--default-character-set}
option when you start the server.  The character sets available depend
on the @code{--with-charset=charset} and @code{--with-extra-charset=
list-of-charset | complex | all} options to @code{configure}, and the
character set configuration files listed in
@file{SHAREDIR/charsets/Index}.  @xref{configure options}.

If you change the character set when running MySQL (which may
also change the sort order), you must run @code{myisamchk -r -q} on all
tables. Otherwise, your indexes may not be ordered correctly.

When a client connects to a MySQL server, the server sends the
default character set in use to the client.  The client will switch to
use this character set for this connection.

One should use @code{mysql_real_escape_string()} when escaping strings
for a SQL query.  @code{mysql_real_escape_string()} is identical to the
old @code{mysql_escape_string()} function, except that it takes the @code{MYSQL}
connection handle as the first parameter.

If the client is compiled with different paths than where the server is
installed and the user who configured MySQL didn't include all
character sets in the MySQL binary, one must specify for
the client where it can find the additional character sets it will need
if the server runs with a different character set than the client.

One can specify this by putting in a MySQL option file:

@example
[client]
character-sets-dir=/usr/local/mysql/share/mysql/charsets
@end example

where the path points to the directory in which the dynamic MySQL character
sets are stored.

One can force the client to use specific character set by specifying:

@example
[client]
default-character-set=character-set-name
@end example

but normally this is never needed.

@menu
* German character set::        German character set
@end menu

@node German character set,  , Character sets, Character sets
@subsubsection German character set

To get German sorting order, you should start @code{mysqld} with
@code{--default-character-set=latin_de}.  This will give you the following
characteristics.

When sorting and comparing string's the following mapping is done on the
strings before doing the comparison:

@example
ä  ->  ae
ö  ->  oe
ü  ->  ue
ß  ->  ss
@end example

All accented characters, are converted to their un-accented uppercase
counterpart.  All letters are converted to uppercase.

When comparing strings with @code{LIKE} the one -> two character mapping
is not done. All letters are converted to uppercase. Accent are removed
from all letters except: @code{Ü}, @code{ü}, @code{Ö}, @code{ö},
@code{Ä} and @code{ä}.


@node Languages, Adding character set, Character sets, Localisation
@subsection Non-English Error Messages

@cindex error messages, languages
@cindex messages, languages
@cindex files, error messages
@cindex language support

@code{mysqld} can issue error messages in the following languages:
Czech, Danish, Dutch, English (the default), Estonian, French, German, Greek,
Hungarian, Italian, Japanese, Korean, Norwegian, Norwegian-ny, Polish,
Portuguese, Romanian, Russian, Slovak, Spanish, and Swedish.

To start @code{mysqld} with a particular language, use either the
@code{--language=lang} or @code{-L lang} options. For example:

@example
shell> mysqld --language=swedish
@end example

or:

@example
shell> mysqld --language=/usr/local/share/swedish
@end example

Note that all language names are specified in lowercase.

The language files are located (by default) in
@file{@var{mysql_base_dir}/share/@var{LANGUAGE}/}.

To update the error message file, you should edit the @file{errmsg.txt} file
and execute the following command to generate the @file{errmsg.sys} file:

@example
shell> comp_err errmsg.txt errmsg.sys
@end example

If you upgrade to a newer version of MySQL, remember to repeat
your changes with the new @file{errmsg.txt} file.


@node Adding character set, Character arrays, Languages, Localisation
@subsection Adding a New Character Set

@cindex character sets, adding
@cindex adding, character sets

To add another character set to MySQL, use the following procedure.

Decide if the set is simple or complex.  If the character set
does not need to use special string collating routines for
sorting and does not need multi-byte character support, it is
simple.  If it needs either of those features, it is complex.

For example, @code{latin1} and @code{danish} are simple charactersets while
@code{big5} or @code{czech} are complex character sets.

In the following section, we have assumed that you name your character
set @code{MYSET}.

For a simple character set do the following:

@enumerate
@item
Add MYSET to the end of the @file{sql/share/charsets/Index} file
Assign an unique number to it.

@item
Create the file @file{sql/share/charsets/MYSET.conf}.
(You can use @file{sql/share/charsets/latin1.conf} as a base for this.)

The syntax for the file very simple:

@itemize @bullet
@item
Comments start with a '#' character and proceed to the end of the line.
@item
Words are separated by arbitrary amounts of whitespace.
@item
When defining the character set, every word must be a number in hexadecimal
format
@item
The @code{ctype} array takes up the first 257 words. The
@code{to_lower}, @code{to_upper} and @code{sort_order} arrays take up
256 words each after that.
@end itemize

@xref{Character arrays}.

@item
Add the character set name to the @code{CHARSETS_AVAILABLE} and
@code{COMPILED_CHARSETS} lists in @code{configure.in}.

@item
Reconfigure, recompile, and test.

@end enumerate

For a complex character set do the following:

@enumerate
@item
Create the file @file{strings/ctype-MYSET.c} in the MySQL source
distribution.

@item
Add MYSET to the end of the @file{sql/share/charsets/Index} file.
Assign an unique number to it.

@item
Look at one of the existing @file{ctype-*.c} files to see what needs to
be defined, for example @file{strings/ctype-big5.c}. Note that the
arrays in your file must have names like @code{ctype_MYSET},
@code{to_lower_MYSET}, and so on.  This corresponds to the arrays
in the simple character set. @xref{Character arrays}.  For a complex
character set

@item
Near the top of the file, place a special comment like this:

@example
/*
 * This comment is parsed by configure to create ctype.c,
 * so don't change it unless you know what you are doing.
 *
 * .configure. number_MYSET=MYNUMBER
 * .configure. strxfrm_multiply_MYSET=N
 * .configure. mbmaxlen_MYSET=N
 */
@end example

The @code{configure} program uses this comment to include
the character set into the MySQL library automatically.

The strxfrm_multiply and mbmaxlen lines will be explained in
the following sections.  Only include these if you need the string
collating functions or the multi-byte character set functions,
respectively.

@item
You should then create some of the following functions:

@itemize @bullet
@item @code{my_strncoll_MYSET()}
@item @code{my_strcoll_MYSET()}
@item @code{my_strxfrm_MYSET()}
@item @code{my_like_range_MYSET()}
@end itemize

@xref{String collating}.

@item
Add the character set name to the @code{CHARSETS_AVAILABLE} and
@code{COMPILED_CHARSETS} lists in @code{configure.in}.

@item
Reconfigure, recompile, and test.
@end enumerate

The file @file{sql/share/charsets/README} includes some more instructions.

If you want to have the character set included in the MySQL
distribution, mail a patch to @email{internals@@lists.mysql.com}.


@node Character arrays, String collating, Adding character set, Localisation
@subsection The Character Definition Arrays

@code{to_lower[]} and @code{to_upper[]} are simple arrays that hold the
lowercase and uppercase characters corresponding to each member of the
character set.  For example:

@example
to_lower['A'] should contain 'a'
to_upper['a'] should contain 'A'
@end example

@code{sort_order[]} is a map indicating how characters should be ordered for
comparison and sorting purposes.  For many character sets, this is the same as
@code{to_upper[]} (which means sorting will be case-insensitive).
MySQL will sort characters based on the value of
@code{sort_order[character]}.  For more complicated sorting rules, see
the discussion of string collating below. @xref{String collating}.

@code{ctype[]} is an array of bit values, with one element for one character.
(Note that @code{to_lower[]}, @code{to_upper[]}, and @code{sort_order[]}
are indexed by character value, but @code{ctype[]} is indexed by character
value + 1.  This is an old legacy to be able to handle @code{EOF}.)

You can find the following bitmask definitions in @file{m_ctype.h}:

@example
#define _U      01      /* Uppercase */
#define _L      02      /* Lowercase */
#define _N      04      /* Numeral (digit) */
#define _S      010     /* Spacing character */
#define _P      020     /* Punctuation */
#define _C      040     /* Control character */
#define _B      0100    /* Blank */
#define _X      0200    /* heXadecimal digit */
@end example

The @code{ctype[]} entry for each character should be the union of the
applicable bitmask values that describe the character.  For example,
@code{'A'} is an uppercase character (@code{_U}) as well as a
hexadecimal digit (@code{_X}), so @code{ctype['A'+1]} should contain the
value:

@example
_U + _X = 01 + 0200 = 0201
@end example


@node String collating, Multi-byte characters, Character arrays, Localisation
@subsection String Collating Support

@cindex collating, strings
@cindex string collating

If the sorting rules for your language are too complex to be handled
with the simple @code{sort_order[]} table, you need to use the string
collating functions.

Right now the best documentation on this is the character sets that are
already implemented.  Look at the @code{big5}, @code{czech}, @code{gbk},
@code{sjis}, and @code{tis160} character sets for examples.

You must specify the @code{strxfrm_multiply_MYSET=N} value in the
special comment at the top of the file.  @code{N} should be set to
the maximum ratio the strings may grow during @code{my_strxfrm_MYSET} (it
must be a positive integer).


@node Multi-byte characters, Problems with character sets, String collating, Localisation
@subsection Multi-byte Character Support

@cindex characters, multi-byte
@cindex multi-byte characters

If your want to add support for a new character set that includes
multi-byte characters, you need to use the multi-byte character
functions.

Right now the best documentation on this is the character sets that are
already implemented.  Look at the euc_kr, gb2312, gbk, sjis and ujis
character sets for examples. These are implemented in the
@code{ctype-'charset'.c} files in the @file{strings} directory.

You must specify the @code{mbmaxlen_MYSET=N} value in the special
comment at the top of the source file.  @code{N} should be set to the
size in bytes of the largest character in the set.

@node Problems with character sets,  , Multi-byte characters, Localisation
@subsection Problems With Character Sets

If you try to use a character set that is not compiled into your binary,
you can run into a couple of different problems:

@itemize @bullet
@item
Your program has a wrong path to where the character sets are stored.
(Default @file{/usr/local/mysql/share/mysql/charsets}).
This can be fixed by using the @code{--character-sets-dir}
option to the program in question.
@item
The character set is a multi-byte-character set that can't be loaded
dynamically.  In this case you have to recompile the program with the
support for the character set.
@item
The character set is a dynamic character set, but you don't have a
configure file for it.  In this case you should install the configure
file for the character set from a new MySQL distribution.
@item
Your @file{Index} file doesn't contain the name for the character set.

@example
ERROR 1105: File '/usr/local/share/mysql/charsets/?.conf' not found
(Errcode: 2)
@end example

In this case you should either get a new @code{Index} file or add
by hand the name of any missing character sets.
@end itemize

For MyISAM tables, you can check the character set name and number for a
table with @code{myisamchk -dvv table_name}.


@node Server-Side Scripts, Client-Side Scripts, Localisation, MySQL Database Administration
@section MySQL Server-Side Scripts and Utilities

@menu
* Server-Side Overview::        Overview of the Server-Side Scripts and Utilities
* safe_mysqld::                 @code{safe_mysqld}, The Wrapper Around @code{mysqld}
* mysqld_multi::                @code{mysqld_multi}, Program for Managing Multiple MySQL Servers
* myisampack::                  @code{myisampack}, The MySQL Compressed Read-only Table Generator
* mysqld-max::                  @code{mysqld-max}, An Extended @code{mysqld} Server
@end menu


@node Server-Side Overview, safe_mysqld, Server-Side Scripts, Server-Side Scripts
@subsection Overview of the Server-Side Scripts and Utilities

@cindex environment variables
@cindex programs, list of

All MySQL clients that communicate with the server using the
@code{mysqlclient} library use the following environment variables:

@tindex @code{MYSQL_UNIX_PORT} environment variable
@tindex environment variable, @code{MYSQL_UNIX_PORT}
@tindex @code{MYSQL_TCP_PORT} environment variable
@tindex environment variable, @code{MYSQL_TCP_PORT}
@tindex @code{MYSQL_PWD} environment variable
@tindex environment variable, @code{MYSQL_PWD}
@tindex @code{MYSQL_DEBUG} environment variable
@tindex environment variable, @code{MYSQL_DEBUG}
@multitable @columnfractions .25 .60
@item @strong{Name} @tab @strong{Description}
@item @code{MYSQL_UNIX_PORT} @tab The default socket; used for connections to @code{localhost}
@item @code{MYSQL_TCP_PORT}  @tab The default TCP/IP port
@item @code{MYSQL_PWD} @tab The default password
@item @code{MYSQL_DEBUG} @tab Debug-trace options when debugging
@item @code{TMPDIR} @tab The directory where temporary tables/files are created
@end multitable

Use of @code{MYSQL_PWD} is insecure.
@xref{Connecting}.

@tindex @code{MYSQL_HISTFILE} environment variable
@tindex environment variable, @code{MYSQL_HISTFILE}
@tindex @code{HOME} environment variable
@tindex environment variable, @code{HOME}
@cindex history file
@cindex command-line history
@tindex .mysql_history file
The @file{mysql} client uses the file named in the @code{MYSQL_HISTFILE}
environment variable to save the command-line history. The default value for
the history file is @file{$HOME/.mysql_history}, where @code{$HOME} is the
value of the @code{HOME} environment variable. @xref{Environment variables}.

All MySQL programs take many different options. However, every
MySQL program provides a @code{--help} option that you can use
to get a full description of the program's different options. For example, try
@code{mysql --help}.

You can override default options for all standard client programs with an
option file. @ref{Option files}.

The following list briefly describes the MySQL programs:

@table @code

@cindex @code{myisamchk}
@item myisamchk
Utility to describe, check, optimise, and repair MySQL tables.
Because @code{myisamchk} has many functions, it is described in its own
chapter. @xref{MySQL Database Administration}.

@cindex @code{make_binary_distribution}
@item make_binary_distribution
Makes a binary release of a compiled MySQL. This could be sent
by FTP to @file{/pub/mysql/Incoming} on @code{support.mysql.com} for the
convenience of other MySQL users.

@cindex @code{msql2mysql}
@item msql2mysql
A shell script that converts @code{mSQL} programs to MySQL. It doesn't
handle all cases, but it gives a good start when converting.

@cindex @code{mysqlaccess}
@item mysqlaccess
A script that checks the access privileges for a host, user, and database
combination.

@cindex @code{mysqladmin}
@item mysqladmin
Utility for performing administrative operations, such as creating or
dropping databases, reloading the grant tables, flushing tables to disk, and
reopening log files.  @code{mysqladmin} can also be used to retrieve version,
process, and status information from the server.
@xref{mysqladmin, , @code{mysqladmin}}.

@cindex @code{mysqlbug}
@item mysqlbug
The MySQL bug report script.  This script should always be used when
filing a bug report to the MySQL list.

@cindex @code{mysqld}
@item mysqld
The SQL daemon. This should always be running.

@cindex @code{mysqldump}
@item mysqldump
Dumps a MySQL database into a file as SQL statements or
as tab-separated text files. Enhanced freeware originally by Igor Romanenko.
@xref{mysqldump, , @code{mysqldump}}.

@cindex @code{mysqlimport}
@item mysqlimport
Imports text files into their respective tables using @code{LOAD DATA
INFILE}. @xref{mysqlimport, , @code{mysqlimport}}.

@cindex @code{mysqlshow}
@item mysqlshow
Displays information about databases, tables, columns, and indexes.

@cindex @code{mysql_install_db}
@item mysql_install_db
Creates the MySQL grant tables with default privileges. This is
usually executed only once, when first installing MySQL
on a system.

@cindex @code{replace}
@item replace
A utility program that is used by @code{msql2mysql}, but that has more
general applicability as well.  @code{replace} changes strings in place in
files or on the standard input. Uses a finite state machine to match longer
strings first. Can be used to swap strings. For example, this command
swaps @code{a} and @code{b} in the given files:

@example
shell> replace a b b a -- file1 file2 ...
@end example
@end table


@node safe_mysqld, mysqld_multi, Server-Side Overview, Server-Side Scripts
@subsection @code{safe_mysqld}, The Wrapper Around @code{mysqld}

@cindex tools, safe_mysqld
@cindex scripts
@cindex @code{safe_mysqld}

@code{safe_mysqld} is the recommended way to start a @code{mysqld}
daemon on Unix.  @code{safe_mysqld} adds some safety features such as
restarting the server when an error occurs and logging run-time
information to a log file.

If you don't use @code{--mysqld=#} or @code{--mysqld-version=#}
@code{safe_mysqld} will use an executable named @code{mysqld-max} if it
exists. If not, @code{safe_mysqld} will start @code{mysqld}.
This makes it very easy to test to use @code{mysqld-max} instead of
@code{mysqld}; just copy @code{mysqld-max} to where you have
@code{mysqld} and it will be used.

Normally one should never edit the @code{safe_mysqld} script, but
instead put the options to @code{safe_mysqld} in the
@code{[safe_mysqld]} section in the @file{my.cnf}
file. @code{safe_mysqld} will read all options from the @code{[mysqld]},
@code{[server]} and @code{[safe_mysqld]} sections from the option files.
@xref{Option files}.

Note that all options on the command-line to @code{safe_mysqld} are passed
to @code{mysqld}.  If you wants to use any options in @code{safe_mysqld} that
@code{mysqld} doesn't support, you must specify these in the option file.

Most of the options to @code{safe_mysqld} are the same as the options to
@code{mysqld}. @xref{Command-line options}.

@code{safe_mysqld} supports the following options:

@table @code
@item --basedir=path
@item --core-file-size=#
Size of the core file @code{mysqld} should be able to create. Passed to @code{ulimit -c}.
@item --datadir=path
@item --defaults-extra-file=path
@item --defaults-file=path
@item --err-log=path
@item --ledir=path
Path to @code{mysqld}
@item --log=path
@item --mysqld=mysqld-version
Name of the @code{mysqld} version in the @code{ledir} directory you want to start.
@item --mysqld-version=version
Similar to @code{--mysqld=} but here you only give the suffix for @code{mysqld}.
For example if you use @code{--mysqld-version=max}, @code{safe_mysqld} will
start the @code{ledir/mysqld-max} version.  If the argument to
@code{--mysqld-version} is empty, @code{ledir/mysqld} will be used.
@item --no-defaults
@item --open-files-limit=#
Number of files @code{mysqld} should be able to open. Passed to @code{ulimit -n}. Note that you need to start @code{safe_mysqld} as root for this to work properly!
@item --pid-file=path
@item --port=#
@item --socket=path
@item --timezone=#
Set the timezone (the @code{TZ}) variable to the value of this parameter.
@item --user=#
@end table

The @code{safe_mysqld} script is written so that it normally is able to start
a server that was installed from either a source or a binary version of
MySQL, even if these install the server in slightly different
locations.  @code{safe_mysqld} expects one of these conditions to be true:

@itemize @bullet
@item
The server and databases can be found relative to the directory from which
@code{safe_mysqld} is invoked.  @code{safe_mysqld} looks under its working
directory for @file{bin} and @file{data} directories (for binary
distributions) or for @file{libexec} and @file{var} directories (for source
distributions).  This condition should be met if you execute
@code{safe_mysqld} from your MySQL installation directory (for
example, @file{/usr/local/mysql} for a binary distribution).

@item
If the server and databases cannot be found relative to the working directory,
@code{safe_mysqld} attempts to locate them by absolute pathnames.  Typical
locations are @file{/usr/local/libexec} and @file{/usr/local/var}.
The actual locations are determined when the distribution was built from which
@code{safe_mysqld} comes.  They should be correct if
MySQL was installed in a standard location.
@end itemize

Because @code{safe_mysqld} will try to find the server and databases relative
to its own working directory, you can install a binary distribution of
MySQL anywhere, as long as you start @code{safe_mysqld} from the
MySQL installation directory:

@example
shell> cd mysql_installation_directory
shell> bin/safe_mysqld &
@end example

If @code{safe_mysqld} fails, even when invoked from the MySQL
installation directory, you can modify it to use the path to @code{mysqld}
and the pathname options that are correct for your system.  Note that if you
upgrade MySQL in the future, your modified version of
@code{safe_mysqld} will be overwritten, so you should make a copy of your
edited version that you can reinstall.


@node mysqld_multi, myisampack, safe_mysqld, Server-Side Scripts
@subsection @code{mysqld_multi}, Program for Managing Multiple MySQL Servers

@cindex tools, mysqld_multi
@cindex scripts
@cindex multi mysqld
@cindex @code{mysqld_multi}

@code{mysqld_multi} is meant for managing several @code{mysqld}
processes running in different Unix sockets and TCP/IP ports.

The program will search for group(s) named [mysqld#] from my.cnf (or the
given --config-file=...), where # can be any positive number starting
from 1. These groups should be the same as the usual @code{[mysqld]}
group (e.g., options to @code{mysqld}, see the MySQL manual for detailed
information about this group), but with those port, socket, etc. options
that are wanted for each separate @code{mysqld} processes. The number in
the group name has another function; it can be used for starting,
stopping, or reporting some specific @code{mysqld} servers with this
program. See the usage and options here for more information.

@example
Usage: mysqld_multi [OPTIONS] @{start|stop|report@} [GNR,GNR,GNR...]
or     mysqld_multi [OPTIONS] @{start|stop|report@} [GNR-GNR,GNR,GNR-GNR,...]
@end example

The GNR above means the group number. You can start, stop or report
any GNR, or several of them at the same time. (See --example) The GNRs
list can be comma separated or combined with a dash, of which the latter
means that all the GNRs between GNR1-GNR2 will be affected. Without
GNR argument all the found groups will be either started, stopped, or
reported. Note that you must not have any white spaces in the GNR
list. Anything after a white space is ignored.

@code{mysqld_multi} supports the following options:

@table @code
@cindex config-file option
@item --config-file=...
Alternative config file. Note: This will not affect this program's own
options (group @code{[mysqld_multi]}), but only groups
[mysqld#]. Without this option everything will be searched from the
ordinary my.cnf file.
@cindex example option
@item --example
Give an example of a config file.
@cindex help option
@item --help
Print this help and exit.
@cindex log option
@item --log=...
Log file. Full path to and the name for the log file. Note: If the file
exists, everything will be appended.
@cindex mysqladmin option
@item --mysqladmin=...
@code{mysqladmin} binary to be used for a server shutdown.
@cindex mysqld option
@item --mysqld=...
@code{mysqld} binary to be used. Note that you can give
@code{safe_mysqld} to this option also. The options are passed to
@code{mysqld}. Just make sure you have @code{mysqld} in your environment
variable @code{PATH} or fix @code{safe_mysqld}.
@cindex no-log option
@item --no-log
Print to stdout instead of the log file. By default the log file is
turned on.
@cindex password option
@item --password=...
Password for user for @code{mysqladmin}.
@cindex tcp-ip option
@item --tcp-ip
Connect to the MySQL server(s) via the TCP/IP port instead of
the Unix socket. This affects stopping and reporting.  If a socket file
is missing, the server may still be running, but can be accessed only
via the TCP/IP port.  By default connecting is done via the Unix socket.
@cindex user option
@item --user=...
MySQL user for @code{mysqladmin}.
@cindex version option
@item --version
Print the version number and exit.
@end table

Some notes about @code{mysqld_multi}:

@itemize @bullet
@item
Make sure that the MySQL user, who is stopping the
@code{mysqld} services (e.g using the @code{mysqladmin}) have the same
password and username for all the data directories accessed (to the
'mysql' database) And make sure that the user has the 'Shutdown_priv'
privilege! If you have many data- directories and many different 'mysql'
databases with different passwords for the MySQL 'root' user,
you may want to create a common 'multi_admin' user for each using the
same password (see below). Example how to do it:
@example
shell> mysql -u root -S /tmp/mysql.sock -proot_password -e
"GRANT SHUTDOWN ON *.* TO multi_admin@@localhost IDENTIFIED BY 'multipass'"
@xref{Privileges}.
@end example
You will have to do the above for each @code{mysqld} running in each
data directory, that you have (just change the socket, -S=...).
@item
@code{pid-file} is very important, if you are using @code{safe_mysqld}
to start @code{mysqld} (e.g., @code{--mysqld=safe_mysqld}) Every @code{mysqld}
should have its own @code{pid-file}. The advantage using
@code{safe_mysqld} instead of @code{mysqld} directly here is, that
@code{safe_mysqld} 'guards' every @code{mysqld} process and will restart
it, if a @code{mysqld} process fails due to signal kill -9, or
similar. (Like segmentation fault, which MySQL should never do,
of course ;) Please note that @code{safe_mysqld} script may require that
you start it from a certain place. This means that you may have to @code{cd} to
a certain directory, before you start the @code{mysqld_multi}. If
you have problems starting, please see the @code{safe_mysqld}
script. Check especially the lines:
@example
--------------------------------------------------------------------------
MY_PWD=`pwd` Check if we are starting this relative (for the binary
release) if test -d /data/mysql -a -f ./share/mysql/english/errmsg.sys
-a -x ./bin/mysqld
--------------------------------------------------------------------------
@xref{safe_mysqld, , @code{safe_mysqld}}.
@end example
The above test should be successful, or you may encounter problems.
@item
Beware of the dangers starting multiple @code{mysqld}s in the same data
directory.  Use separate data directories, unless you @strong{know} what
you are doing!
@item
The socket file and the TCP/IP port must be different for every @code{mysqld}.
@item
The first and fifth @code{mysqld} group were intentionally left out from
the example.  You may have 'gaps' in the config file. This gives you
more flexibility.  The order in which the @code{mysqlds} are started or
stopped depends on the order in which they appear in the config file.
@item
When you want to refer to a certain group using GNR with this program,
just use the number in the end of the group name ([mysqld# <== ).
@item
You may want to use option '--user' for @code{mysqld}, but in order to
do this you need to be root when you start the @code{mysqld_multi}
script. Having the option in the config file doesn't matter; you will
just get a warning, if you are not the superuser and the @code{mysqlds}
are started under @strong{your} Unix account. @strong{Important}: Make
sure that the @code{pid-file} and the data directory are
read+write(+execute for the latter one) accessible for @strong{that}
Unix user, who the specific @code{mysqld} process is started
as. @strong{Do not} use the Unix root account for this, unless you
@strong{know} what you are doing!
@item
@strong{Most important}: Make sure that you understand the meanings of
the options that are passed to the @code{mysqld}s and @strong{why one
would want} to have separate @code{mysqld} processes. Starting multiple
@code{mysqld}s in one data directory @strong{will not} give you extra
performance in a threaded system!
@end itemize

@xref{Multiple servers}.

This is an example of the config file on behalf of @code{mysqld_multi}.

@example
# This file should probably be in your home dir (~/.my.cnf) or /etc/my.cnf
# Version 2.1 by Jani Tolonen

[mysqld_multi]
mysqld     = /usr/local/bin/safe_mysqld
mysqladmin = /usr/local/bin/mysqladmin
user       = multi_admin
password   = multipass

[mysqld2]
socket     = /tmp/mysql.sock2
port       = 3307
pid-file   = /usr/local/mysql/var2/hostname.pid2
datadir    = /usr/local/mysql/var2
language   = /usr/local/share/mysql/english
user       = john

[mysqld3]
socket     = /tmp/mysql.sock3
port       = 3308
pid-file   = /usr/local/mysql/var3/hostname.pid3
datadir    = /usr/local/mysql/var3
language   = /usr/local/share/mysql/swedish
user       = monty

[mysqld4]
socket     = /tmp/mysql.sock4
port       = 3309
pid-file   = /usr/local/mysql/var4/hostname.pid4
datadir    = /usr/local/mysql/var4
language   = /usr/local/share/mysql/estonia
user       = tonu

[mysqld6]
socket     = /tmp/mysql.sock6
port       = 3311
pid-file   = /usr/local/mysql/var6/hostname.pid6
datadir    = /usr/local/mysql/var6
language   = /usr/local/share/mysql/japanese
user       = jani
@end example

@xref{Option files}.


@node myisampack, mysqld-max, mysqld_multi, Server-Side Scripts
@subsection @code{myisampack}, The MySQL Compressed Read-only Table Generator

@cindex compressed tables
@cindex tables, compressed
@cindex MyISAM, compressed tables
@cindex @code{myisampack}
@cindex @code{pack_isam}

@code{myisampack} is used to compress MyISAM tables, and @code{pack_isam}
is used to compress ISAM tables. Because ISAM tables are deprecated, we
will only discuss @code{myisampack} here, but everything said about
@code{myisampack} should also be true for @code{pack_isam}.

@code{myisampack} works by compressing each column in the table separately.
The information needed to decompress columns is read into memory when the
table is opened. This results in much better performance when accessing
individual records, because you only have to uncompress exactly one record, not
a much larger disk block as when using Stacker on MS-DOS.
Usually, @code{myisampack} packs the datafile 40%-70%.

MySQL uses memory mapping (@code{mmap()}) on compressed tables and
falls back to normal read/write file usage if @code{mmap()} doesn't work.

Please note the following:
@itemize @bullet
@item
After packing, the table is read-only. This is generally intended
(such as when accessing packed tables on a CD). Also allowing writes
to a packed table is on our TODO list but with low priority.
@item
@code{myisampack} can also pack @code{BLOB} or @code{TEXT} columns.
The older @code{pack_isam} (for @code{ISAM} tables) can not do this.
@end itemize

@code{myisampack} is invoked like this:

@example
shell> myisampack [options] filename ...
@end example

Each filename should be the name of an index (@file{.MYI}) file.  If you
are not in the database directory, you should specify the pathname to the
file.  It is permissible to omit the @file{.MYI} extension.

@code{myisampack} supports the following options:

@table @code
@item -b, --backup
Make a backup of the table as @code{tbl_name.OLD}.

@item -#, --debug=debug_options
Output debug log. The @code{debug_options} string often is
@code{'d:t:o,filename'}.

@item -f, --force
Force packing of the table even if it becomes bigger or if the temporary file
exists.  @code{myisampack} creates a temporary file named @file{tbl_name.TMD}
while it compresses the table.  If you kill @code{myisampack}, the @file{.TMD}
file may not be deleted.  Normally, @code{myisampack} exits with an error if
it finds that @file{tbl_name.TMD} exists.  With @code{--force},
@code{myisampack} packs the table anyway.

@item -?, --help
Display a help message and exit.

@item -j big_tbl_name, --join=big_tbl_name
Join all tables named on the command-line into a single table
@code{big_tbl_name}.  All tables that are to be combined
@strong{must} be identical (same column names and types, same indexes, etc.).

@item -p #, --packlength=#
Specify the record length storage size, in bytes.  The value should be 1, 2,
or 3.  (@code{myisampack} stores all rows with length pointers of 1, 2, or 3
bytes.  In most normal cases, @code{myisampack} can determine the right length
value before it begins packing the file, but it may notice during the packing
process that it could have used a shorter length. In this case,
@code{myisampack} will print a note that the next time you pack the same file,
you could use a shorter record length.)

@item -s, --silent
Silent mode.  Write output only when errors occur.

@item -t, --test
Don't actually pack table, just test packing it.

@item -T dir_name, --tmp_dir=dir_name
Use the named directory as the location in which to write the temporary table.

@item -v, --verbose
Verbose mode.  Write information about progress and packing result.

@item -V, --version
Display version information and exit.

@item -w, --wait

Wait and retry if table is in use.  If the @code{mysqld} server was
invoked with the @code{--skip-locking} option, it is not a good idea to
invoke @code{myisampack} if the table might be updated during the
packing process.
@end table

@cindex examples, compressed tables
The sequence of commands shown here illustrates a typical table compression
session:

@example
shell> ls -l station.*
-rw-rw-r--   1 monty    my         994128 Apr 17 19:00 station.MYD
-rw-rw-r--   1 monty    my          53248 Apr 17 19:00 station.MYI
-rw-rw-r--   1 monty    my           5767 Apr 17 19:00 station.frm

shell> myisamchk -dvv station

MyISAM file:     station
Isam-version:  2
Creation time: 1996-03-13 10:08:58
Recover time:  1997-02-02  3:06:43
Data records:              1192  Deleted blocks:              0
Datafile: Parts:           1192  Deleted data:                0
Datafile pointer (bytes):     2  Keyfile pointer (bytes):     2
Max datafile length:   54657023  Max keyfile length:   33554431
Recordlength:               834
Record format: Fixed length

table description:
Key Start Len Index   Type                       Root  Blocksize    Rec/key
1   2     4   unique  unsigned long              1024       1024          1
2   32    30  multip. text                      10240       1024          1

Field Start Length Type
1     1     1
2     2     4
3     6     4
4     10    1
5     11    20
6     31    1
7     32    30
8     62    35
9     97    35
10    132   35
11    167   4
12    171   16
13    187   35
14    222   4
15    226   16
16    242   20
17    262   20
18    282   20
19    302   30
20    332   4
21    336   4
22    340   1
23    341   8
24    349   8
25    357   8
26    365   2
27    367   2
28    369   4
29    373   4
30    377   1
31    378   2
32    380   8
33    388   4
34    392   4
35    396   4
36    400   4
37    404   1
38    405   4
39    409   4
40    413   4
41    417   4
42    421   4
43    425   4
44    429   20
45    449   30
46    479   1
47    480   1
48    481   79
49    560   79
50    639   79
51    718   79
52    797   8
53    805   1
54    806   1
55    807   20
56    827   4
57    831   4

shell> myisampack station.MYI
Compressing station.MYI: (1192 records)
- Calculating statistics

normal:     20  empty-space:      16  empty-zero:        12  empty-fill:  11
pre-space:   0  end-space:        12  table-lookups:      5  zero:         7
Original trees:  57  After join: 17
- Compressing file
87.14%

shell> ls -l station.*
-rw-rw-r--   1 monty    my         127874 Apr 17 19:00 station.MYD
-rw-rw-r--   1 monty    my          55296 Apr 17 19:04 station.MYI
-rw-rw-r--   1 monty    my           5767 Apr 17 19:00 station.frm

shell> myisamchk -dvv station

MyISAM file:     station
Isam-version:  2
Creation time: 1996-03-13 10:08:58
Recover time:  1997-04-17 19:04:26
Data records:              1192  Deleted blocks:              0
Datafile: Parts:           1192  Deleted data:                0
Datafilepointer (bytes):      3  Keyfile pointer (bytes):     1
Max datafile length:   16777215  Max keyfile length:     131071
Recordlength:               834
Record format: Compressed

table description:
Key Start Len Index   Type                       Root  Blocksize    Rec/key
1   2     4   unique  unsigned long             10240       1024          1
2   32    30  multip. text                      54272       1024          1

Field Start Length Type                         Huff tree  Bits
1     1     1      constant                             1     0
2     2     4      zerofill(1)                          2     9
3     6     4      no zeros, zerofill(1)                2     9
4     10    1                                           3     9
5     11    20     table-lookup                         4     0
6     31    1                                           3     9
7     32    30     no endspace, not_always              5     9
8     62    35     no endspace, not_always, no empty    6     9
9     97    35     no empty                             7     9
10    132   35     no endspace, not_always, no empty    6     9
11    167   4      zerofill(1)                          2     9
12    171   16     no endspace, not_always, no empty    5     9
13    187   35     no endspace, not_always, no empty    6     9
14    222   4      zerofill(1)                          2     9
15    226   16     no endspace, not_always, no empty    5     9
16    242   20     no endspace, not_always              8     9
17    262   20     no endspace, no empty                8     9
18    282   20     no endspace, no empty                5     9
19    302   30     no endspace, no empty                6     9
20    332   4      always zero                          2     9
21    336   4      always zero                          2     9
22    340   1                                           3     9
23    341   8      table-lookup                         9     0
24    349   8      table-lookup                        10     0
25    357   8      always zero                          2     9
26    365   2                                           2     9
27    367   2      no zeros, zerofill(1)                2     9
28    369   4      no zeros, zerofill(1)                2     9
29    373   4      table-lookup                        11     0
30    377   1                                           3     9
31    378   2      no zeros, zerofill(1)                2     9
32    380   8      no zeros                             2     9
33    388   4      always zero                          2     9
34    392   4      table-lookup                        12     0
35    396   4      no zeros, zerofill(1)               13     9
36    400   4      no zeros, zerofill(1)                2     9
37    404   1                                           2     9
38    405   4      no zeros                             2     9
39    409   4      always zero                          2     9
40    413   4      no zeros                             2     9
41    417   4      always zero                          2     9
42    421   4      no zeros                             2     9
43    425   4      always zero                          2     9
44    429   20     no empty                             3     9
45    449   30     no empty                             3     9
46    479   1                                          14     4
47    480   1                                          14     4
48    481   79     no endspace, no empty               15     9
49    560   79     no empty                             2     9
50    639   79     no empty                             2     9
51    718   79     no endspace                         16     9
52    797   8      no empty                             2     9
53    805   1                                          17     1
54    806   1                                           3     9
55    807   20     no empty                             3     9
56    827   4      no zeros, zerofill(2)                2     9
57    831   4      no zeros, zerofill(1)                2     9
@end example

The information printed by @code{myisampack} is described here:

@table @code
@item normal
The number of columns for which no extra packing is used.

@item empty-space
The number of columns containing
values that are only spaces; these will occupy 1 bit.

@item empty-zero
The number of columns containing
values that are only binary 0's; these will occupy 1 bit.

@item empty-fill
The number of integer columns that don't occupy the full byte range of their
type; these are changed to a smaller type (for example, an @code{INTEGER}
column may be changed to @code{MEDIUMINT}).

@item pre-space
The number of decimal columns that are stored with leading spaces. In this
case, each value will contain a count for the number of leading spaces.

@item end-space
The number of columns that have a lot of trailing spaces.  In this case, each
value will contain a count for the number of trailing spaces.

@item table-lookup
The column had only a small number of different values, which were
converted to an @code{ENUM} before Huffman compression.

@item zero
The number of columns for which all values are zero.

@item Original trees
The initial number of Huffman trees.

@item After join
The number of distinct Huffman trees left after joining
trees to save some header space.
@end table

After a table has been compressed, @code{myisamchk -dvv} prints additional
information about each field:

@table @code
@item Type
The field type may contain the following descriptors:

@table @code
@item constant
All rows have the same value.

@item no endspace
Don't store endspace.

@item no endspace, not_always
Don't store endspace and don't do end space compression for all values.

@item no endspace, no empty
Don't store endspace. Don't store empty values.

@item table-lookup
The column was converted to an @code{ENUM}.

@item zerofill(n)
The most significant @code{n} bytes in the value are always 0 and are not
stored.

@item no zeros
Don't store zeros.

@item always zero
0 values are stored in 1 bit.
@end table

@item Huff tree
The Huffman tree associated with the field.

@item Bits
The number of bits used in the Huffman tree.
@end table

After you have run @code{pack_isam}/@code{myisampack} you must run
@code{isamchk}/@code{myisamchk} to re-create the index.  At this time you
can also sort the index blocks and create statistics needed for
the MySQL optimiser to work more efficiently:

@example
myisamchk -rq --analyze --sort-index table_name.MYI
isamchk   -rq --analyze --sort-index table_name.ISM
@end example

After you have installed the packed table into the MySQL database
directory you should do @code{mysqladmin flush-tables} to force @code{mysqld}
to start using the new table.

If you want to unpack a packed table, you can do this with the
@code{--unpack} option to @code{isamchk} or @code{myisamchk}.


@node mysqld-max,  , myisampack, Server-Side Scripts
@subsection @code{mysqld-max}, An Extended @code{mysqld} Server

@cindex @code{mysqld-max}

@code{mysqld-max} is the MySQL server (@code{mysqld}) configured with
the following configure options:

@multitable @columnfractions .30 .45
@item @strong{Option} @tab @strong{Comment}
@item --with-server-suffix=-max @tab Add a suffix to the @code{mysqld} version string.
@item --with-innodb @tab Support for InnoDB tables.
@item --with-bdb @tab Support for Berkeley DB (BDB) tables
@item CFLAGS=-DUSE_SYMDIR @tab Symbolic links support for Windows.
@end multitable

You can find the MySQL-max binaries at
@uref{http://www.mysql.com/downloads/mysql-max-3.23.html}.

The Windows MySQL binary distributions includes both the
standard @code{mysqld.exe} binary and the @code{mysqld-max.exe} binary.
@uref{http://www.mysql.com/downloads/mysql-3.23.html}.
@xref{Windows installation}.

Note that as InnoDB and Berkeley DB are not available for all platforms,
some of the @code{Max} binaries may not have support for both of these.
You can check which table types are supported by doing the following
query:

@example
mysql> SHOW VARIABLES LIKE "have_%";
+---------------+-------+
| Variable_name | Value |
+---------------+-------+
| have_bdb      | YES   |
| have_innodb   | NO    |
| have_isam     | YES   |
| have_raid     | NO    |
| have_openssl  | NO    |
+---------------+-------+
@end example

The meaning of the values are:

@multitable @columnfractions .15 .85
@item @strong{Value} @tab @strong{Meaning}
@item @code{YES} @tab The option is activated and usable.
@item @code{NO} @tab MySQL is not compiled with support for this option.
@item @code{DISABLED} @tab The xxxx option is disabled because one started @code{mysqld} with @code{--skip-xxxx} or because one didn't start @code{mysqld} with all needed options to enable the option.  In this case the @code{hostname.err} file should contain a reason for why the option is disabled.
@end multitable

@strong{Note}:  To be able to create InnoDB tables you @strong{must} edit
your startup options to include at least the @code{innodb_data_file_path}
option. @xref{InnoDB start}.

To get better performance for BDB tables, you should add some configuration
options for these too. @xref{BDB start}.

@code{safe_mysqld} will automatically try to start any @code{mysqld} binary
with the @code{-max} prefix. This makes it very easy to test out a
another @code{mysqld} binary in an existing installation.  Just
run @code{configure} with the options you want and then install the
new @code{mysqld} binary as @code{mysqld-max} in the same directory
where your old @code{mysqld} binary is. @xref{safe_mysqld, , @code{safe_mysqld}}.

The @code{mysqld-max} RPM uses the above mentioned @code{safe_mysqld}
feature. It just installs the @code{mysqld-max} executable and
@code{safe_mysqld} will automatically use this executable when
@code{safe_mysqld} is restarted.

The following table shows which table types our standard @strong{MySQL-Max}
binaries includes:

@multitable @columnfractions .15 .10 .10
@item @strong{System} @tab @code{BDB} @tab @code{InnoDB}
@item AIX 4.3 @tab N @tab Y
@item HP-UX 11.0 @tab N @tab Y
@item Linux-Alpha @tab N	@tab Y
@item Linux-Intel @tab Y @tab Y
@item Linux-IA64 @tab N @tab Y
@item Solaris-Intel @tab N @tab Y
@item Solaris-SPARC @tab Y @tab Y
@item Caldera (SCO) OSR5 @tab Y @tab Y
@item UnixWare @tab Y @tab Y
@item Windows/NT @tab Y @tab Y
@end multitable


@node Client-Side Scripts, Log Files, Server-Side Scripts, MySQL Database Administration
@section MySQL Client-Side Scripts and Utilities

@menu
* Client-Side Overview::        Overview of the Client-Side Scripts and Utilities
* mysql::                       The Command-line Tool
* mysqladmin::                  @code{mysqladmin}, Administrating a MySQL Server
* Using mysqlcheck::            Using @code{mysqlcheck} for Table Maintenance and Crash Recovery
* mysqldump::                   @code{mysqldump}, Dumping Table Structure and Data
* mysqlhotcopy::                @code{mysqlhotcopy}, Copying MySQL Databases and Tables
* mysqlimport::                 @code{mysqlimport}, Importing Data from Text Files
* mysqlshow::                   Showing Databases, Tables, and Columns
* perror::                      @code{perror}, Explaining Error Codes
* Batch Commands::              How to Run SQL Commands from a Text File
@end menu


@node Client-Side Overview, mysql, Client-Side Scripts, Client-Side Scripts
@subsection Overview of the Client-Side Scripts and Utilities

@cindex environment variables
@cindex programs, list of

All MySQL clients that communicate with the server using the
@code{mysqlclient} library use the following environment variables:

@tindex @code{MYSQL_UNIX_PORT} environment variable
@tindex environment variable, @code{MYSQL_UNIX_PORT}
@tindex @code{MYSQL_TCP_PORT} environment variable
@tindex environment variable, @code{MYSQL_TCP_PORT}
@tindex @code{MYSQL_PWD} environment variable
@tindex environment variable, @code{MYSQL_PWD}
@tindex @code{MYSQL_DEBUG} environment variable
@tindex environment variable, @code{MYSQL_DEBUG}
@multitable @columnfractions .25 .60
@item @strong{Name} @tab @strong{Description}
@item @code{MYSQL_UNIX_PORT} @tab The default socket; used for connections to @code{localhost}
@item @code{MYSQL_TCP_PORT}  @tab The default TCP/IP port
@item @code{MYSQL_PWD} @tab The default password
@item @code{MYSQL_DEBUG} @tab Debug-trace options when debugging
@item @code{TMPDIR} @tab The directory where temporary tables/files are created
@end multitable

Use of @code{MYSQL_PWD} is insecure.
@xref{Connecting}.

@tindex @code{MYSQL_HISTFILE} environment variable
@tindex environment variable, @code{MYSQL_HISTFILE}
@tindex @code{HOME} environment variable
@tindex environment variable, @code{HOME}
@cindex history file
@cindex command-line history
@tindex .mysql_history file
The @file{mysql} client uses the file named in the @code{MYSQL_HISTFILE}
environment variable to save the command-line history. The default value for
the history file is @file{$HOME/.mysql_history}, where @code{$HOME} is the
value of the @code{HOME} environment variable. @xref{Environment variables}.

All MySQL programs take many different options. However, every
MySQL program provides a @code{--help} option that you can use
to get a full description of the program's different options. For example, try
@code{mysql --help}.

You can override default options for all standard client programs with an
option file. @ref{Option files}.

The following list briefly describes the MySQL programs:

@table @code

@cindex @code{myisamchk}
@item myisamchk
Utility to describe, check, optimise, and repair MySQL tables.
Because @code{myisamchk} has many functions, it is described in its own
chapter. @xref{MySQL Database Administration}.

@cindex @code{make_binary_distribution}
@item make_binary_distribution
Makes a binary release of a compiled MySQL. This could be sent
by FTP to @file{/pub/mysql/Incoming} on @code{support.mysql.com} for the
convenience of other MySQL users.

@cindex @code{msql2mysql}
@item msql2mysql
A shell script that converts @code{mSQL} programs to MySQL. It doesn't
handle all cases, but it gives a good start when converting.

@cindex @code{mysqlaccess}
@item mysqlaccess
A script that checks the access privileges for a host, user, and database
combination.

@cindex @code{mysqladmin}
@item mysqladmin
Utility for performing administrative operations, such as creating or
dropping databases, reloading the grant tables, flushing tables to disk, and
reopening log files.  @code{mysqladmin} can also be used to retrieve version,
process, and status information from the server.
@xref{mysqladmin, , @code{mysqladmin}}.

@cindex @code{mysqlbug}
@item mysqlbug
The MySQL bug report script.  This script should always be used when
filing a bug report to the MySQL list.

@cindex @code{mysqld}
@item mysqld
The SQL daemon. This should always be running.

@cindex @code{mysqldump}
@item mysqldump
Dumps a MySQL database into a file as SQL statements or
as tab-separated text files. Enhanced freeware originally by Igor Romanenko.
@xref{mysqldump, , @code{mysqldump}}.

@cindex @code{mysqlimport}
@item mysqlimport
Imports text files into their respective tables using @code{LOAD DATA
INFILE}. @xref{mysqlimport, , @code{mysqlimport}}.

@cindex @code{mysqlshow}
@item mysqlshow
Displays information about databases, tables, columns, and indexes.

@cindex @code{mysql_install_db}
@item mysql_install_db
Creates the MySQL grant tables with default privileges. This is
usually executed only once, when first installing MySQL
on a system.

@cindex @code{replace}
@item replace
A utility program that is used by @code{msql2mysql}, but that has more
general applicability as well.  @code{replace} changes strings in place in
files or on the standard input. Uses a finite state machine to match longer
strings first. Can be used to swap strings. For example, this command
swaps @code{a} and @code{b} in the given files:

@example
shell> replace a b b a -- file1 file2 ...
@end example
@end table


@node mysql, mysqladmin, Client-Side Overview, Client-Side Scripts
@subsection @code{mysql}, The Command-line Tool

@cindex command-line tool
@cindex tools, command-line
@cindex scripts
@cindex @code{mysql}

@code{mysql} is a simple SQL shell (with GNU @code{readline} capabilities).
It supports interactive and non-interactive use. When used interactively,
query results are presented in an ASCII-table format. When used
non-interactively (for example, as a filter), the result is presented in
tab-separated format.  (The output format can be changed using command-line
options.)  You can run scripts simply like this:

@example
shell> mysql database < script.sql > output.tab
@end example

If you have problems due to insufficient memory in the client, use the
@code{--quick} option!  This forces @code{mysql} to use
@code{mysql_use_result()} rather than @code{mysql_store_result()} to
retrieve the result set.

Using @code{mysql} is very easy. Just start it as follows:
@code{mysql database} or @code{mysql --user=user_name --password=your_password database}. Type a SQL statement, end it with @samp{;}, @samp{\g}, or @samp{\G}
and press Enter.

@cindex @code{mysql} command-line options
@cindex command-line options, @code{mysql}
@cindex options, command-line, @code{mysql}
@cindex startup parameters, @code{mysql}
@code{mysql} supports the following options:

@table @code
@cindex @code{help}, @code{mysql} option
@item -?, --help
Display this help and exit.

@cindex @code{no-auto-rehash}, @code{mysql} option
@item -A, --no-auto-rehash
No automatic rehashing. One has to use 'rehash' to get table and field
completion. This gives a quicker start of mysql.

@cindex @code{prompt}, @code{mysql} option
@item --prompt=...
Set the mysql prompt to specified format.

@cindex @code{no-beep}, @code{mysql} option
@item -b, --no-beep
Turn off beep-on-error.

@cindex @code{batch}, @code{mysql} option
@item -B, --batch
Print results with a tab as separator, each row on a new line. Doesn't use
history file.

@cindex @code{character-sets-dir}, @code{mysql} option
@item --character-sets-dir=...
Directory where character sets are located.

@cindex @code{compress}, @code{mysql} option
@item -C, --compress
Use compression in server/client protocol.

@cindex @code{debug}, @code{mysql} option
@item -#, --debug[=...]
Debug log. Default is 'd:t:o,/tmp/mysql.trace'.

@cindex @code{database}, @code{mysql} option
@item -D, --database=...
Database to use. This is mainly useful in the @file{my.cnf} file.

@cindex @code{default-character-set}, @code{mysql} option
@item --default-character-set=...
Set the default character set.

@cindex @code{execute}, @code{mysql} option
@item -e, --execute=...
Execute command and quit. (Output like with --batch)

@cindex @code{vertical}, @code{mysql} option
@item -E, --vertical
Print the output of a query (rows) vertically. Without this option you
can also force this output by ending your statements with @code{\G}.

@cindex @code{force}, @code{mysql} option
@item -f, --force
Continue even if we get a SQL error.

@cindex @code{no-named-commands}, @code{mysql} option
@item -g, --no-named-commands
Named commands are disabled. Use \* form only, or use named commands
only in the beginning of a line ending with a semicolon (@samp{;}). Since
Version 10.9, the client now starts with this option @strong{enabled} by default!
With the -g option, long format commands will still work from the first
line, however.

@cindex @code{enable-named-commands}, @code{mysql} option
@item -G, --enable-named-commands
Named commands are @strong{enabled}.  Long format commands are allowed as
well as shortened \* commands.

@cindex @code{ignore-space}, @code{mysql} option
@item -i, --ignore-space
Ignore space after function names.

@cindex @code{host}, @code{mysql} option
@item -h, --host=...
Connect to the given host.

@cindex @code{html}, @code{mysql} option
@item -H, --html
Produce HTML output.

@cindex @code{skip-line-numbers}, @code{mysql} option
@item -L, --skip-line-numbers
Don't write line number for errors. Useful when one wants to compare result
files that includes error messages

@cindex @code{no-pager}, @code{mysql} option
@item --no-pager
Disable pager and print to stdout. See interactive help (\h) also.

@cindex @code{no-tee}, @code{mysql} option
@item --no-tee
Disable outfile. See interactive help (\h) also.

@cindex @code{unbuffered}, @code{mysql} option
@item -n, --unbuffered
Flush buffer after each query.

@cindex @code{skip-column-names}, @code{mysql} option
@item -N, --skip-column-names
Don't write column names in results.

@cindex @code{set-variable}, @code{mysql} option
@item -O, --set-variable var=option
Give a variable a value. @code{--help} lists variables.

@cindex @code{one-database}, @code{mysql} option
@item -o, --one-database
Only update the default database. This is useful for skipping updates to
other database in the update log.

@cindex @code{pager}, @code{mysql} option
@item @code{--pager[=...]}
Output type. Default is your @code{ENV} variable @code{PAGER}. Valid
pagers are less, more, cat [> filename], etc.  See interactive help (\h)
also. This option does not work in batch mode. Pager works only in Unix.

@cindex @code{password}, @code{mysql} option
@item -p[password], --password[=...]
Password to use when connecting to server. If a password is not given on
the command-line, you will be prompted for it.  Note that if you use the
short form @code{-p} you can't have a space between the option and the
password.

@cindex @code{port}, @code{mysql} option
@item -P  --port=...
TCP/IP port number to use for connection.

@cindex @code{quick}, @code{mysql} option
@item -q, --quick
Don't cache result, print it row-by-row. This may slow down the server
if the output is suspended. Doesn't use history file.

@cindex @code{raw}, @code{mysql} option
@item -r, --raw
Write column values without escape conversion. Used with @code{--batch}

@cindex @code{silent}, @code{mysql} option
@item -s, --silent
Be more silent.

@cindex @code{socket}, @code{mysql} option
@item -S  --socket=...
Socket file to use for connection.

@cindex @code{table}, @code{mysql} option
@item -t  --table
Output in table format. This is default in non-batch mode.

@cindex @code{debug-info}, @code{mysql} option
@item -T, --debug-info
Print some debug information at exit.

@cindex @code{tee}, @code{mysql} option
@item --tee=...
Append everything into outfile. See interactive help (\h) also. Does not
work in batch mode.

@cindex @code{user}, @code{mysql} option
@item -u, --user=#
User for login if not current user.

@cindex @code{safe-updates}, @code{mysql} option
@item -U, --safe-updates[=#], --i-am-a-dummy[=#]
Only allow @code{UPDATE} and @code{DELETE} that uses keys. See below for
more information about this option.  You can reset this option if you have
it in your @file{my.cnf} file by using @code{--safe-updates=0}.

@cindex @code{verbose}, @code{mysql} option
@item -v, --verbose
More verbose output (-v -v -v gives the table output format).

@cindex @code{version}, @code{mysql} option
@item -V, --version
Output version information and exit.

@cindex @code{wait}, @code{mysql} option
@item -w, --wait
Wait and retry if connection is down instead of aborting.
@end table

You can also set the following variables with @code{-O} or
@code{--set-variable}:

@cindex timeout, @code{connect_timeout} variable
@cindex @code{connect_timeout} variable
@cindex @code{max_allowed_packet}
@cindex @code{net_buffer_length}
@cindex @code{select_limit}
@cindex @code{max_join_size}
@multitable @columnfractions .25 .10 .65
@item @strong{Variable Name} @tab @strong{Default} @tab @strong{Description}
@item connect_timeout    @tab 0         @tab Number of seconds before timeout connection.
@item max_allowed_packet @tab 16777216  @tab Max packetlength to send/receive from to server
@item net_buffer_length  @tab 16384     @tab Buffer for TCP/IP and socket communication
@item select_limit       @tab 1000      @tab Automatic limit for SELECT when using --i-am-a-dummy
@item max_join_size      @tab 1000000   @tab Automatic limit for rows in a join when using --i-am-a-dummy.
@end multitable

If you type 'help' on the command-line, @code{mysql} will print out the
commands that it supports:

@cindex commands, list of
@example
mysql> help

MySQL commands:
help    (\h)    Display this text.
?       (\h)    Synonym for `help'.
clear   (\c)    Clear command.
connect (\r)    Reconnect to the server.
                Optional arguments are db and host.
edit    (\e)    Edit command with $EDITOR.
ego     (\G)    Send command to mysql server,
                display result vertically.
exit    (\q)    Exit mysql. Same as quit.
go      (\g)    Send command to mysql server.
nopager (\n)    Disable pager, print to stdout.
notee   (\t)    Don't write into outfile.
pager   (\P)    Set PAGER [to_pager].
                Print the query results via PAGER.
print   (\p)    Print current command.
prompt  (\R)    Change your mysql prompt.
quit    (\q)    Quit mysql.
rehash  (\#)    Rebuild completion hash.
source  (\.)    Execute a SQL script file.
                Takes a file name as an argument.
status  (\s)    Get status information from the server.
tee     (\T)    Set outfile [to_outfile].
                Append everything into given outfile.
use     (\u)    Use another database.
                Takes database name as argument.
@end example

The @code{pager} command works only in Unix.

@cindex @code{status} command
The @code{status} command gives you some information about the
connection and the server you are using. If you are running in the
@code{--safe-updates} mode, @code{status} will also print the values for
the @code{mysql} variables that affect your queries.

@cindex @code{safe-mode} command
A useful startup option for beginners (introduced in MySQL
Version 3.23.11) is @code{--safe-updates} (or @code{--i-am-a-dummy} for
users that has at some time done a @code{DELETE FROM table_name} but
forgot the @code{WHERE} clause).  When using this option, @code{mysql}
sends the following command to the MySQL server when opening
the connection:

@example
SET SQL_SAFE_UPDATES=1,SQL_SELECT_LIMIT=#select_limit#,
    SQL_MAX_JOIN_SIZE=#max_join_size#"
@end example

where @code{#select_limit#} and @code{#max_join_size#} are variables that
can be set from the @code{mysql} command-line. @xref{SET OPTION, @code{SET}}.

The effect of the above is:

@itemize @bullet
@item
You are not allowed to do an @code{UPDATE} or @code{DELETE} statement
if you don't have a key constraint in the @code{WHERE} part. One can,
however, force an @code{UPDATE/DELETE} by using @code{LIMIT}:
@example
UPDATE table_name SET not_key_column=# WHERE not_key_column=# LIMIT 1;
@end example
@item
All big results are automatically limited to @code{#select_limit#} rows.
@item
@code{SELECT}'s that will probably need to examine more than
@code{#max_join_size} row combinations will be aborted.
@end itemize

Some useful hints about the @code{mysql} client:

Some data is much more readable when displayed vertically, instead of
the usual horizontal box type output. For example longer text, which
includes new lines, is often much easier to be read with vertical
output.

@example
mysql> SELECT * FROM mails WHERE LENGTH(txt) < 300 lIMIT 300,1\G
*************************** 1. row ***************************
  msg_nro: 3068
     date: 2000-03-01 23:29:50
time_zone: +0200
mail_from: Monty
    reply: monty@@no.spam.com
  mail_to: "Thimble Smith" <tim@@no.spam.com>
      sbj: UTF-8
      txt: >>>>> "Thimble" == Thimble Smith writes:

Thimble> Hi.  I think this is a good idea.  Is anyone familiar with UTF-8
Thimble> or Unicode? Otherwise, I'll put this on my TODO list and see what
Thimble> happens.

Yes, please do that.

Regards,
Monty
     file: inbox-jani-1
     hash: 190402944
1 row in set (0.09 sec)
@end example

@itemize @bullet
@item
For logging, you can use the @code{tee} option. The @code{tee} can be
started with option @code{--tee=...}, or from the command-line
interactively with command @code{tee}. All the data displayed on the
screen will also be appended into a given file. This can be very useful
for debugging purposes also. The @code{tee} can be disabled from the
command-line with command @code{notee}. Executing @code{tee} again
starts logging again. Without a parameter the previous file will be
used. Note that @code{tee} will flush the results into the file after
each command, just before the command-line appears again waiting for the
next command.

@item
Browsing, or searching the results in the interactive mode in Unix less,
more, or any other similar program, is now possible with option
@code{--pager[=...]}. Without argument, @code{mysql} client will look
for environment variable PAGER and set @code{pager} to that.
@code{pager} can be started from the interactive command-line with
command @code{pager} and disabled with command @code{nopager}.  The
command takes an argument optionally and the @code{pager} will be set to
that. Command @code{pager} can be called without an argument, but this
requires that the option @code{--pager} was used, or the @code{pager}
will default to stdout. @code{pager} works only in Unix, since it uses
the popen() function, which doesn't exist in Windows. In Windows, the
@code{tee} option can be used instead, although it may not be as handy
as @code{pager} can be in some situations.

@item
A few tips about @code{pager}: You can use it to write to a file:
@example
mysql> pager cat > /tmp/log.txt
@end example
and the results will only go to a file. You can also pass any options
for the programs that you want to use with the @code{pager}:
@example
mysql> pager less -n -i -S
@end example
From the above do note the option '-S'. You may find it very useful when
browsing the results; try the option with horizontal output (end
commands with '\g', or ';') and with vertical output (end commands with
'\G'). Sometimes a very wide result set is hard to be read from the screen,
with option -S to less you can browse the results within the interactive
less from left to right, preventing lines longer than your screen from
being continued to the next line. This can make the result set much more
readable. You can swith the mode between on and off within the interactive
less with '-S'. See the 'h' for more help about less.

@item
Last (unless you already understood this from the above examples ;) you
can combine very complex ways to handle the results, for example the
following would send the results to two files in two different
directories, on two different hard-disks mounted on /dr1 and /dr2, yet
let the results still be seen on the screen via less:
@example
mysql> pager cat | tee /dr1/tmp/res.txt | \
tee /dr2/tmp/res2.txt | less -n -i -S
@end example

@item
You can also combine the two functions above; have the @code{tee}
enabled, @code{pager} set to 'less' and you will be able to browse the
results in unix 'less' and still have everything appended into a file
the same time. The difference between Unix @code{tee} used with the
@code{pager} and the @code{mysql} client in-built @code{tee}, is that
the in-built @code{tee} works even if you don't have the Unix @code{tee}
available. The in-built @code{tee} also logs everything that is printed
on the screen, where the Unix @code{tee} used with @code{pager} doesn't
log quite that much. Last, but not least, the interactive @code{tee} is
more handy to switch on and off, when you want to log something into a
file, but want to be able to turn the feature off sometimes.
@end itemize

@cindex @code{prompt} command
You can change the prompt in the @code{mysql} command-line client.

You can use the following prompt options:
@c FIX these columnfractions have NOT been measured!
@multitable @columnfractions .10 .50
@item @strong{Option} @tab @strong{Description}
@item \v   @tab  mysqld version
@item \d   @tab  database in use
@item \h   @tab  host connected to
@item \p   @tab  port connected on
@item \u   @tab  username
@item \U   @tab  full username@@host
@item \\   @tab  @samp{\}
@item \n   @tab  new line break
@item \t   @tab  tab
@item \    @tab  space
@item \_   @tab  space
@item \R   @tab  military hour time (0-23)
@item \r   @tab  standard hour time (1-12)
@item \m   @tab  minutes
@item \y   @tab  two digit year 
@item \Y   @tab  four digit year
@item \D   @tab  full date format
@item \s   @tab  seconds
@item \w   @tab  day of the week in three letter format (Mon, Tue, ...)
@item \P   @tab  am/pm
@item \o   @tab  month in number format
@item \O   @tab  month in three letter format (Jan, Feb, ...)
@item \c   @tab  counter that counts up for each command you do
@end multitable

@samp{\} followed by any other letter just becomes that letter.

You may set the prompt in the following places:

@table @strong
@item Environment Variable
You may set the @code{MYSQL_PS1} environment variable to a prompt string.  For
example:

@example
shell> export MYSQL_PS1="(\u@@\h) [\d]> "
@end example

@item @file{my.cnf}
@item @file{.my.cnf}
You may set the @code{prompt} option in any MySQL configuration file, in the
@code{mysql} group.  For example:

@example
[mysql]
prompt=(\u@@\h) [\d]>\_
@end example

@item Command Line
You may set the @code{--prompt} option on the command line to @code{mysql}.
For example:

@example
shell> mysql --prompt="(\u@@\h) [\d]> "

(user@@host) [database]> 
@end example

@item Interactively
You may also use the @code{prompt} (or @code{\R}) command to change your 
prompt interactively.  For example:

@example
mysql> prompt (\u@@\h) [\d]>\_
PROMPT set to '(\u@@\h) [\d]>\_'
(user@@host) [database]> 
(user@@host) [database]> prompt
Returning to default PROMPT of mysql> 
mysql> 
@end example

@end table


@node mysqladmin, Using mysqlcheck, mysql, Client-Side Scripts
@subsection @code{mysqladmin}, Administrating a MySQL Server

@cindex administration, server
@cindex server administration
@cindex @code{mysladmn}

A utility for performing administrative operations. The syntax is:

@example
shell> mysqladmin [OPTIONS] command [command-option] command ...
@end example

You can get a list of the options your version of @code{mysqladmin} supports
by executing @code{mysqladmin --help}.

The current @code{mysqladmin} supports the following commands:

@table @code
@item create databasename
Create a new database.

@item drop databasename
Delete a database and all its tables.

@item extended-status
Gives an extended status message from the server.

@item flush-hosts
Flush all cached hosts.

@item flush-logs
Flush all logs.

@item flush-tables
Flush all tables.

@item flush-privileges
Reload grant tables (same as reload).

@item kill id,id,...
Kill mysql threads.

@item password
Set a new password. Change old password to new-password.

@item ping
Check if mysqld is alive.

@item processlist
Show list of active threads in server.

@item reload
Reload grant tables.

@item refresh
Flush all tables and close and open logfiles.

@item shutdown
Take server down.

@item slave-start
Start slave replication thread.

@item slave-stop
Stop slave replication thread.

@item status
Gives a short status message from the server.

@item variables
Prints variables available.

@item version
Get version info from server.
@end table

All commands can be shortened to their unique prefix.  For example:

@example
shell> mysqladmin proc stat
+----+-------+-----------+----+-------------+------+-------+------+
| Id | User  | Host      | db | Command     | Time | State | Info |
+----+-------+-----------+----+-------------+------+-------+------+
| 6  | monty | localhost |    | Processlist | 0    |       |      |
+----+-------+-----------+----+-------------+------+-------+------+
Uptime: 10077  Threads: 1  Questions: 9  Slow queries: 0
Opens: 6 Flush tables: 1  Open tables: 2
Memory in use: 1092K  Max memory used: 1116K
@end example

@cindex status command, results
The @code{mysqladmin status} command result has the following columns:

@cindex uptime
@multitable @columnfractions .20 .65
@item @strong{Column} @tab @strong{Description}
@item Uptime @tab Number of seconds the MySQL server has been up.
@cindex threads
@item Threads @tab Number of active threads (clients).
@cindex questions
@item Questions @tab Number of questions from clients since @code{mysqld} was started.
@cindex slow queries
@item Slow queries @tab Queries that have taken more than @code{long_query_time} seconds. @xref{Slow query log}.
@cindex opens
@item Opens @tab How many tables @code{mysqld} has opened.
@cindex flush tables
@cindex tables, flush
@item Flush tables @tab Number of @code{flush ...}, @code{refresh}, and @code{reload} commands.
@cindex open tables
@item Open tables @tab Number of tables that are open now.
@cindex memory use
@item Memory in use @tab Memory allocated directly by the @code{mysqld} code (only available when MySQL is compiled with --with-debug=full).
@cindex max memory used
@item Max memory used @tab Maximum memory allocated directly by the @code{mysqld} code (only available when MySQL is compiled with --with-debug=full).
@end multitable

If you do @code{myslqadmin shutdown} on a socket (in other words, on a
the computer where @code{mysqld} is running), @code{mysqladmin} will
wait until the MySQL @code{pid-file} is removed to ensure that
the @code{mysqld} server has stopped properly.


@node Using mysqlcheck, mysqldump, mysqladmin, Client-Side Scripts
@subsection Using @code{mysqlcheck} for Table Maintenance and Crash Recovery

Since MySQL version 3.23.38 you will be able to use a new
checking and repairing tool for @code{MyISAM} tables. The difference to
@code{myisamchk} is that @code{mysqlcheck} should be used when the
@code{mysqld} server is running, where as @code{myisamchk} should be used
when it is not. The benefit is that you no longer have to take the
server down for checking or repairing your tables.

@code{mysqlcheck} uses MySQL server commands @code{CHECK},
@code{REPAIR}, @code{ANALYZE} and @code{OPTIMIZE} in a convenient way
for the user.

There are three alternative ways to invoke @code{mysqlcheck}:

@example
shell> mysqlcheck [OPTIONS] database [tables]
shell> mysqlcheck [OPTIONS] --databases DB1 [DB2 DB3...]
shell> mysqlcheck [OPTIONS] --all-databases
@end example

So it can be used in a similar way as @code{mysqldump} when it
comes to what databases and tables you want to choose.

@code{mysqlcheck} does have a special feature compared to the other
clients; the default behavior, checking tables (-c), can be changed by
renaming the binary. So if you want to have a tool that repairs tables
by default, you should just copy @code{mysqlcheck} to your harddrive
with a new name, @code{mysqlrepair}, or alternatively make a symbolic
link to @code{mysqlrepair} and name the symbolic link as
@code{mysqlrepair}. If you invoke @code{mysqlrepair} now, it will repair
tables by default.

The names that you can use to change @code{mysqlcheck} default behavior
are here:

@example
mysqlrepair:   The default option will be -r
mysqlanalyze:  The default option will be -a
mysqloptimize: The default option will be -o
@end example

The options available for @code{mysqlcheck} are listed here, please
check what your version supports with @code{mysqlcheck --help}.

@table @code
@item -A, --all-databases
Check all the databases. This will be same as --databases with all
databases selected
@item -1, --all-in-1
Instead of making one query for each table, execute all queries in 1
query separately for each database. Table names will be in a comma
separated list.
@item  -a, --analyze
Analyse given tables.
@item --auto-repair
If a checked table is corrupted, automatically fix it. Repairing will be
done after all tables have been checked, if corrupted ones were found.
@item -#, --debug=...
Output debug log. Often this is 'd:t:o,filename'
@item --character-sets-dir=...
Directory where character sets are
@item -c, --check
Check table for errors
@item -C, --check-only-changed
Check only tables that have changed since last check or haven't been
closed properly.
@item --compress
Use compression in server/client protocol.
@item -?, --help
Display this help message and exit.
@item -B, --databases
To check several databases. Note the difference in usage; in this case
no tables are given. All name arguments are regarded as database names.
@item --default-character-set=...
Set the default character set
@item -F, --fast
Check only tables that hasn't been closed properly
@item -f, --force
Continue even if we get an sql-error.
@item -e, --extended
If you are using this option with CHECK TABLE, it will ensure that the
table is 100 percent consistent, but will take a long time.

If you are using this option with REPAIR TABLE, it will run an extended
repair on the table, which may not only take a long time to execute, but
may produce a lot of garbage rows also!
@item -h, --host=...
Connect to host.
@item -m, --medium-check
Faster than extended-check, but only finds 99.99 percent of all
errors. Should be good enough for most cases.
@item -o, --optimize
Optimise table
@item -p, --password[=...]
Password to use when connecting to server. If password is not given
it's solicited on the tty.
@item -P, --port=...
Port number to use for connection.
@item -q, --quick
If you are using this option with CHECK TABLE, it prevents the check
from scanning the rows to check for wrong links. This is the fastest
check.

If you are using this option with REPAIR TABLE, it will try to repair
only the index tree. This is the fastest repair method for a table.
@item -r, --repair
Can fix almost anything except unique keys that aren't unique.
@item -s, --silent
Print only error messages.
@item -S, --socket=...
Socket file to use for connection.
@item --tables
Overrides option --databases (-B).
@item -u, --user=#
User for login if not current user.
@item -v, --verbose
Print info about the various stages.
@item -V, --version
Output version information and exit.
@end table


@node mysqldump, mysqlhotcopy, Using mysqlcheck, Client-Side Scripts
@subsection @code{mysqldump}, Dumping Table Structure and Data

@cindex dumping, databases
@cindex databases, dumping
@cindex tables, dumping
@cindex backing up, databases

@cindex @code{mysqldump}
Utility to dump a database or a collection of database for backup or for
transferring the data to another SQL server (not necessarily a MySQL
server).  The dump will contain SQL statements to create the table
and/or populate the table.

If you are doing a backup on the server, you should consider using
the @code{mysqlhotcopy} instead. @xref{mysqlhotcopy, , @code{mysqlhotcopy}}.

@example
shell> mysqldump [OPTIONS] database [tables]
OR     mysqldump [OPTIONS] --databases [OPTIONS] DB1 [DB2 DB3...]
OR     mysqldump [OPTIONS] --all-databases [OPTIONS]
@end example

If you don't give any tables or use the @code{--databases} or
@code{--all-databases}, the whole database(s) will be dumped.

You can get a list of the options your version of @code{mysqldump} supports
by executing @code{mysqldump --help}.

Note that if you run @code{mysqldump} without @code{--quick} or
@code{--opt}, @code{mysqldump} will load the whole result set into
memory before dumping the result.  This will probably be a problem if
you are dumping a big database.

Note that if you are using a new copy of the @code{mysqldump} program
and you are going to do a dump that will be read into a very old MySQL
server, you should not use the @code{--opt} or @code{-e} options.

@code{mysqldump} supports the following options:

@table @code
@item --add-locks
Add @code{LOCK TABLES} before and @code{UNLOCK TABLE} after each table dump.
(To get faster inserts into MySQL.)
@item --add-drop-table
Add a @code{drop table} before each create statement.
@item -A, --all-databases
Dump all the databases. This will be same as @code{--databases} with all
databases selected.
@item -a, --all
Include all MySQL-specific create options.
@item --allow-keywords
Allow creation of column names that are keywords.  This works by
prefixing each column name with the table name.
@item -c, --complete-insert
Use complete insert statements (with column names).
@item -C, --compress
Compress all information between the client and the server if both support
compression.
@item -B, --databases
To dump several databases. Note the difference in usage. In this case
no tables are given. All name arguments are regarded as database names.
@code{USE db_name;} will be included in the output before each new database.
@item --delayed
Insert rows with the @code{INSERT DELAYED} command.
@item -e, --extended-insert
Use the new multiline @code{INSERT} syntax. (Gives more compact and
faster inserts statements.)
@item -#, --debug[=option_string]
Trace usage of the program (for debugging).
@item --help
Display a help message and exit.
@item --fields-terminated-by=...
@itemx --fields-enclosed-by=...
@itemx --fields-optionally-enclosed-by=...
@itemx --fields-escaped-by=...
@itemx --lines-terminated-by=...
These options are used with the @code{-T} option and have the same
meaning as the corresponding clauses for @code{LOAD DATA INFILE}.
@xref{LOAD DATA, , @code{LOAD DATA}}.
@item -F, --flush-logs
Flush log file in the MySQL server before starting the dump.
@item -f, --force,
Continue even if we get a SQL error during a table dump.
@item -h, --host=..
Dump data from the MySQL server on the named host. The default host
is @code{localhost}.
@item -l, --lock-tables.
Lock all tables before starting the dump.  The tables are locked with
@code{READ LOCAL} to allow concurrent inserts in the case of @code{MyISAM}
tables.
@item -K, --disable-keys
@code{/*!40000 ALTER TABLE tb_name DISABLE KEYS */;} and
@code{/*!40000 ALTER TABLE tb_name ENABLE KEYS */;}
will be put in the output.  This will make loading the data into a MySQL
4.0 server faster as the indexes are created after all data are inserted.
@item -n, --no-create-db
@code{CREATE DATABASE /*!32312 IF NOT EXISTS*/ db_name;} will not be put in the
output. The above line will be added otherwise, if --databases or
--all-databases option was given.
@item -t, --no-create-info
Don't write table creation information (the @code{CREATE TABLE} statement).
@item -d, --no-data
Don't write any row information for the table.  This is very useful if you
just want to get a dump of the structure for a table!
@item --opt
Same as @code{--quick --add-drop-table --add-locks --extended-insert
--lock-tables}.  Should give you the fastest possible dump for reading
into a MySQL server.
@item -pyour_pass, --password[=your_pass]
The password to use when connecting to the server. If you specify
no @samp{=your_pass} part,
@code{mysqldump} you will be prompted for a password.
@item -P port_num, --port=port_num
The TCP/IP port number to use for connecting to a host.  (This is used for
connections to hosts other than @code{localhost}, for which Unix sockets are
used.)
@item  -q, --quick
Don't buffer query, dump directly to stdout. Uses @code{mysql_use_result()}
to do this.
@item -Q, --quote-names
Quote table and column names within @samp{`} characters.
@item -r, --result-file=...
Direct output to a given file. This option should be used in MSDOS,
because it prevents new line '\n' from being converted to '\n\r' (new
line + carriage return).
@item -S /path/to/socket, --socket=/path/to/socket
The socket file to use when connecting to @code{localhost} (which is the
default host).
@item --tables
Overrides option --databases (-B).
@item  -T, --tab=path-to-some-directory
Creates a @code{table_name.sql} file, that contains the SQL CREATE commands,
and a @code{table_name.txt} file, that contains the data, for each give table.
@strong{Note}: This only works if @code{mysqldump} is run on the same
machine as the @code{mysqld} daemon.  The format of the @file{.txt} file
is made according to the @code{--fields-xxx} and @code{--lines--xxx} options.
@item -u user_name, --user=user_name
The MySQL user name to use when connecting to the server. The
default value is your Unix login name.
@item -O var=option, --set-variable var=option
Set the value of a variable.  The possible variables are listed below.
@item -v, --verbose
Verbose mode.  Print out more information on what the program does.
@item -V, --version
Print version information and exit.
@item -w, --where='where-condition'
Dump only selected records. Note that quotes are mandatory:
@item -X, --xml
Dumps a database as well formed XML
@item -x, --first-slave
Locks all tables across all databases.

@example
"--where=user='jimf'" "-wuserid>1" "-wuserid<1"
@end example
@item -O net_buffer_length=#, where # < 16M
When creating multi-row-insert statements (as with option
@code{--extended-insert} or @code{--opt}), @code{mysqldump} will create
rows up to @code{net_buffer_length} length. If you increase this
variable, you should also ensure that the @code{max_allowed_packet}
variable in the MySQL server is bigger than the
@code{net_buffer_length}.
@end table

The most normal use of @code{mysqldump} is probably for making a backup of
whole databases. @xref{Backup}.

@example
mysqldump --opt database > backup-file.sql
@end example

You can read this back into MySQL with:

@example
mysql database < backup-file.sql
@end example

or

@example
mysql -e "source /patch-to-backup/backup-file.sql" database
@end example

However, it's also very useful to populate another MySQL server with
information from a database:

@example
mysqldump --opt database | mysql ---host=remote-host -C database
@end example

It is possible to dump several databases with one command:

@example
mysqldump --databases database1 [database2 ...] > my_databases.sql
@end example

If all the databases are wanted, one can use:

@example
mysqldump --all-databases > all_databases.sql
@end example


@node mysqlhotcopy, mysqlimport, mysqldump, Client-Side Scripts
@subsection @code{mysqlhotcopy}, Copying MySQL Databases and Tables

@cindex dumping, databases
@cindex databases, dumping
@cindex tables, dumping
@cindex backing up, databases

@code{mysqlhotcopy} is a Perl script that uses @code{LOCK TABLES},
@code{FLUSH TABLES} and @code{cp} or @code{scp} to quickly make a backup
of a database.  It's the fastest way to make a backup of the database,
of single tables but it can only be run on the same machine where the
database directories are.

@example
mysqlhotcopy db_name [/path/to/new_directory]

mysqlhotcopy db_name_1 ... db_name_n /path/to/new_directory

mysqlhotcopy db_name./regex/
@end example

@code{mysqlhotcopy} supports the following options:

@table @code
@item -?, --help
Display a help screen and exit
@item -u, --user=#
User for database login
@item -p, --password=#
Password to use when connecting to server
@item -P, --port=#
Port to use when connecting to local server
@item -S, --socket=#
Socket to use when connecting to local server
@item --allowold
Don't abort if target already exists (rename it _old)
@item --keepold
Don't delete previous (now renamed) target when done
@item --noindices
Don't include full index files in copy to make the backup smaller and faster
The indexes can later be reconstructed with @code{myisamchk -rq.}.
@item --method=#
Method for copy (@code{cp} or @code{scp}).
@item -q, --quiet
Be silent except for errors
@item --debug
Enable debug
@item -n, --dryrun
Report actions without doing them
@item --regexp=#
Copy all databases with names matching regexp
@item --suffix=#
Suffix for names of copied databases
@item --checkpoint=#
Insert checkpoint entry into specified db.table
@item --flushlog
Flush logs once all tables are locked.
@item --tmpdir=#
Temporary directory (instead of /tmp).
@end table

You can use @code{perldoc mysqlhotcopy} to get a more complete
documentation for @code{mysqlhotcopy}.

@code{mysqlhotcopy} reads the groups @code{[client]} and  @code{[mysqlhotcopy]}
from the option files.

To be able to execute @code{mysqlhotcopy} you need write access to the
backup directory, the @strong{select} privilege for the tables you are about to
copy and the MySQL @strong{reload} privilege (to be able to
execute @code{FLUSH TABLES}).


@node mysqlimport, mysqlshow, mysqlhotcopy, Client-Side Scripts
@subsection @code{mysqlimport}, Importing Data from Text Files

@cindex importing, data
@cindex data, importing
@cindex files, text
@cindex text files, importing
@cindex @code{mysqlimport}

@code{mysqlimport} provides a command-line interface to the @code{LOAD DATA
INFILE} SQL statement.  Most options to @code{mysqlimport} correspond
directly to the same options to @code{LOAD DATA INFILE}.
@xref{LOAD DATA, , @code{LOAD DATA}}.

@code{mysqlimport} is invoked like this:

@example
shell> mysqlimport [options] database textfile1 [textfile2 ...]
@end example

For each text file named on the command-line,
@code{mysqlimport} strips any extension from the filename and uses the result
to determine which table to import the file's contents into.  For example,
files named @file{patient.txt}, @file{patient.text}, and @file{patient} would
all be imported into a table named @code{patient}.

@code{mysqlimport} supports the following options:

@table @code
@item -c, --columns=...
This option takes a comma-separated list of field names as an argument.
The field list is used to create a proper @code{LOAD DATA INFILE} command,
which is then passed to MySQL. @xref{LOAD DATA, , @code{LOAD DATA}}.

@item -C, --compress
Compress all information between the client and the server if both support
compression.

@item -#, --debug[=option_string]
Trace usage of the program (for debugging).

@item -d, --delete
Empty the table before importing the text file.

@item --fields-terminated-by=...
@itemx --fields-enclosed-by=...
@itemx --fields-optionally-enclosed-by=...
@itemx --fields-escaped-by=...
@itemx --lines-terminated-by=...
These options have the same meaning as the corresponding clauses for
@code{LOAD DATA INFILE}. @xref{LOAD DATA, , @code{LOAD DATA}}.

@item -f, --force
Ignore errors.  For example, if a table for a text file doesn't exist,
continue processing any remaining files.  Without @code{--force},
@code{mysqlimport} exits if a table doesn't exist.

@item --help
Display a help message and exit.

@item -h host_name, --host=host_name
Import data to the MySQL server on the named host. The default host
is @code{localhost}.

@item -i, --ignore
See the description for the @code{--replace} option.

@item -l, --lock-tables
Lock @strong{all} tables for writing before processing any text files. This
ensures that all tables are synchronised on the server.

@item -L, --local
Read input files from the client.  By default, text files are assumed to be on
the server if you connect to @code{localhost} (which is the default host).

@item -pyour_pass, --password[=your_pass]
The password to use when connecting to the server. If you specify
no @samp{=your_pass} part,
@code{mysqlimport} you will be prompted for a password.

@item -P port_num, --port=port_num
The TCP/IP port number to use for connecting to a host.  (This is used for
connections to hosts other than @code{localhost}, for which Unix sockets are
used.)

@item -r, --replace
The @code{--replace} and @code{--ignore} options control handling of input
records that duplicate existing records on unique key values.  If you specify
@code{--replace}, new rows replace existing rows that have the same unique key
value. If you specify @code{--ignore}, input rows that duplicate an existing
row on a unique key value are skipped.  If you don't specify either option, an
error occurs when a duplicate key value is found, and the rest of the text
file is ignored.

@item  -s, --silent
Silent mode.  Write output only when errors occur.

@item -S /path/to/socket, --socket=/path/to/socket
The socket file to use when connecting to @code{localhost} (which is the
default host).

@item -u user_name, --user=user_name
The MySQL user name to use when connecting to the server. The
default value is your Unix login name.

@item -v, --verbose
Verbose mode.  Print out more information what the program does.

@item -V, --version
Print version information and exit.
@end table

Here is a sample run using @code{mysqlimport}:

@example
$ mysql --version
mysql  Ver 9.33 Distrib 3.22.25, for pc-linux-gnu (i686)
$ uname -a
Linux xxx.com 2.2.5-15 #1 Mon Apr 19 22:21:09 EDT 1999 i586 unknown
$ mysql -e 'CREATE TABLE imptest(id INT, n VARCHAR(30))' test
$ ed
a
100     Max Sydow
101     Count Dracula
.
w imptest.txt
32
q
$ od -c imptest.txt
0000000   1   0   0  \t   M   a   x       S   y   d   o   w  \n   1   0
0000020   1  \t   C   o   u   n   t       D   r   a   c   u   l   a  \n
0000040
$ mysqlimport --local test imptest.txt
test.imptest: Records: 2  Deleted: 0  Skipped: 0  Warnings: 0
$ mysql -e 'SELECT * FROM imptest' test
+------+---------------+
| id   | n             |
+------+---------------+
|  100 | Max Sydow     |
|  101 | Count Dracula |
+------+---------------+
@end example


@node mysqlshow, perror, mysqlimport, Client-Side Scripts
@subsection Showing Databases, Tables, and Columns

@cindex databases, displaying
@cindex displaying, database information
@cindex tables, displaying
@cindex columns, displaying
@cindex showing, database information

@code{mysqlshow} can be used to quickly look at which databases exist,
their tables, and the table's columns.

With the @code{mysql} program you can get the same information with the
@code{SHOW} commands.  @xref{SHOW}.

@code{mysqlshow} is invoked like this:

@example
shell> mysqlshow [OPTIONS] [database [table [column]]]
@end example

@itemize @bullet
@item
If no database is given, all matching databases are shown.
@item
If no table is given, all matching tables in the database are shown.
@item
If no column is given, all matching columns and column types in the table
are shown.
@end itemize

Note that in newer MySQL versions, you only see those
database/tables/columns for which you have some privileges.

If the last argument contains a shell or SQL wildcard (@code{*}, @code{?},
@code{%} or @code{_}) then only what's matched by the wildcard is shown.
This may cause some confusion when you try to display the columns for a
table with a @code{_} as in this case @code{mysqlshow} only shows you
the table names that match the pattern.  This is easily fixed by
adding an extra @code{%} last on the command-line (as a separate
argument).


@node perror, Batch Commands, mysqlshow, Client-Side Scripts
@subsection perror, Explaining Error Codes

@cindex error messages, displaying
@cindex perror

@cindex errno
@cindex Errcode

For most system errors MySQL will, in addition to a internal text message,
also print the system error code in one of the following styles:
@code{message ... (errno: #)} or @code{message ... (Errcode: #)}.

You can find out what the error code means by either examining the
documentation for your system or use the @code{perror} utility.

@code{perror} prints a description for a system error code, or an MyISAM/ISAM
table handler error code.

@code{perror} is invoked like this:

@example
shell> perror [OPTIONS] [ERRORCODE [ERRORCODE...]]

Example:

shell> perror 13 64
Error code  13:  Permission denied
Error code  64:  Machine is not on the network
@end example

Note that the error messages are mostly system dependent!


@node Batch Commands,  , perror, Client-Side Scripts
@subsection How to Run SQL Commands from a Text File

The @code{mysql} client typically is used interactively, like this:

@example
shell> mysql database
@end example

However, it's also possible to put your SQL commands in a file and tell
@code{mysql} to read its input from that file.  To do so, create a text
file @file{text_file} that contains the commands you wish to execute.
Then invoke @code{mysql} as shown here:

@example
shell> mysql database < text_file
@end example

You can also start your text file with a @code{USE db_name} statement.  In
this case, it is unnecessary to specify the database name on the command
line:

@example
shell> mysql < text_file
@end example

If you are already running @code{mysql}, you can execute a SQL
script file using the @code{source} command:

@example
mysql> source filename;
@end example

For more information about batch mode, @ref{Batch mode}.


@node Log Files, Replication, Client-Side Scripts, MySQL Database Administration
@section The MySQL Log Files

@cindex Log files

MySQL has several different log files that can help you find
out what's going on inside @code{mysqld}:

@multitable @columnfractions .20 .80
@item @strong{Log file} @tab @strong{Description}
@item The error log @tab Problems encountering starting, running or stopping @code{mysqld}.
@item The isam log @tab Logs all changes to the ISAM tables. Used only for debugging the isam code.
@item The query log @tab Established connections and executed queries.
@item The update log @tab Deprecated: Stores all statements that changes data
@item The binary log @tab Stores all statements that changes something. Used also for replication
@item The slow log @tab Stores all queries that took more than @code{long_query_time} to execute or didn't use indexes.
@end multitable

All logs can be found in the @code{mysqld} data directory.  You can
force @code{mysqld} to reopen the log files (or in some cases
switch to a new log) by executing @code{FLUSH LOGS}. @xref{FLUSH}.

@menu
* Error log::                   The Error Log
* Query log::                   The General Query Log
* Update log::                  The Update Log
* Binary log::                  The Binary Update Log
* Slow query log::              The Slow Query Log
* Log file maintenance::        Log File Maintenance
@end menu


@node Error log, Query log, Log Files, Log Files
@subsection The Error Log

@code{mysqld} writes all errors to the stderr, which the
@code{safe_mysqld} script redirects to a file called
@code{'hostname'.err}.  (On Windows, @code{mysqld} writes this directly
to @file{\mysql\data\mysql.err}.)

This contains information indicating when @code{mysqld} was started and
stopped and also any critical errors found when running.  If @code{mysqld}
dies unexpectedly and @code{safe_mysqld} needs to restart @code{mysqld},
@code{safe_mysqld} will write a @code{restarted mysqld} row in this
file.  This log also holds a warning if @code{mysqld} notices a table
that needs to be automatically checked or repaired.

On some operating systems, the error log will contain a stack trace
for where @code{mysqld} died. This can be used to find out where
@code{mysqld} died.  @xref{Using stack trace}.


@node Query log, Update log, Error log, Log Files
@subsection The General Query Log

@cindex query log
@cindex files, query log

If you want to know what happens within @code{mysqld}, you should start
it with @code{--log[=file]}.  This will log all connections and queries
to the log file (by default named @file{'hostname'.log}).  This log can
be very useful when you suspect an error in a client and want to know
exactly what @code{mysqld} thought the client sent to it.

By default, the @code{mysql.server} script starts the MySQL
server with the @code{-l} option.  If you need better performance when
you start using MySQL in a production environment, you can
remove the @code{-l} option from @code{mysql.server} or change it to
@code{--log-bin}.

The entries in this log are written as @code{mysqld} receives the questions.
This may be different from the order in which the statements are executed.
This is in contrast to the update log and the binary log which are written
after the query is executed, but before any locks are released.


@node Update log, Binary log, Query log, Log Files
@subsection The Update Log

@cindex update log
@cindex files, update log

@strong{Note}: the update log is replaced by the binary
log. @xref{Binary log}.  With this you can do anything that you can do
with the update log.

When started with the @code{--log-update[=file_name]} option,
@code{mysqld} writes a log file containing all SQL commands that update
data. If no filename is given, it defaults to the name of the host
machine. If a filename is given, but it doesn't contain a path, the file
is written in the data directory. If @file{file_name} doesn't have an
extension, @code{mysqld} will create log file names like so:
@file{file_name.###}, where @code{###} is a number that is incremented each
time you execute @code{mysqladmin refresh}, execute @code{mysqladmin
flush-logs}, execute the @code{FLUSH LOGS} statement, or restart the server.

@strong{Note}: for the above scheme to work, you must not create
your own files with the same filename as the update log + some extensions
that may be regarded as a number, in the directory used by the update log!

If you use the @code{--log} or @code{-l} options, @code{mysqld} writes a
general log with a filename of @file{hostname.log}, and restarts and
refreshes do not cause a new log file to be generated (although it is closed
and reopened). In this case you can copy it (on Unix) by doing:

@example
mv hostname.log hostname-old.log
mysqladmin flush-logs
cp hostname-old.log to-backup-directory
rm hostname-old.log
@end example

Update logging is smart because it logs only statements that really update
data. So an @code{UPDATE} or a @code{DELETE} with a @code{WHERE} that finds no
rows is not written to the log. It even skips @code{UPDATE} statements that
set a column to the value it already has.

The update logging is done immediately after a query completes but before
any locks are released or any commit is done. This ensures that the log
will be logged in the execution order.

If you want to update a database from update log files, you could do the
following (assuming your update logs have names of the form
@file{file_name.###}):

@example
shell> ls -1 -t -r file_name.[0-9]* | xargs cat | mysql
@end example

@code{ls} is used to get all the log files in the right order.

This can be useful if you have to revert to backup files after a crash
and you want to redo the updates that occurred between the time of the backup
and the crash.


@node Binary log, Slow query log, Update log, Log Files
@subsection The Binary Update Log

@cindex binary log
@cindex files, binary log

The intention is that the binary log should replace the update log, so
we recommend you to switch to this log format as soon as possible!

The binary log contains all information that is available in the update
log in a more efficient format. It also contains information about how long
every query that updated the database took.

The binary log is also used when you are replicating a slave from a master.
@xref{Replication}.

When started with the @code{--log-bin[=file_name]} option, @code{mysqld}
writes a log file containing all SQL commands that update data. If no
file name is given, it defaults to the name of the host machine followed
by @code{-bin}. If file name is given, but it doesn't contain a path, the
file is written in the data directory.

If you supply an extension to @code{--log-bin=filename.extension}, the
extension will be silenty removed.

To the binary log filename @code{mysqld} will append an extension that
is a number that is incremented each time you execute @code{mysqladmin
refresh}, execute @code{mysqladmin flush-logs}, execute the @code{FLUSH
LOGS} statement or restart the server. A new binary log will also
automatically be created when it reaches @code{max_bin_log_size}. You can
delete all not active binary log files with the @code{RESET MASTER}
command. @xref{RESET}.

You can use the following options to @code{mysqld} to affect what is logged
to the binary log:

@multitable @columnfractions .38 .62
@item @strong{Option} @tab @strong{Description}

@item @code{binlog-do-db=database_name} @tab
Tells the master it should log updates for the specified database, and
exclude all others not explicitly mentioned.
(Example: @code{binlog-do-db=some_database})

@item @code{binlog-ignore-db=database_name} @tab
Tells the master that updates to the given database should not be logged
to the binary log (Example: @code{binlog-ignore-db=some_database})
@end multitable

To be able to know which different binary log files have been used,
@code{mysqld} will also create a binary log index file that
contains the name of all used binary log files. By default this has the
same name as the binary log file, with the extension @code{'.index'}.
You can change the name of the binary log index file with the
@code{--log-bin-index=[filename]} option.

If you are using replication, you should not delete old binary log
files until you are sure that no slave will ever need to use them.
One way to do this is to do @code{mysqladmin flush-logs} once a day and then
remove any logs that are more than 3 days old.

You can examine the binary log file with the @code{mysqlbinlog} command.
For example, you can update a MySQL server from the binary log
as follows:

@example
mysqlbinlog log-file | mysql -h server_name
@end example

You can also use the @code{mysqlbinlog} program to read the binary log
directly from a remote MySQL server!

@code{mysqlbinlog --help} will give you more information of how to use
this program!

If you are using @code{BEGIN [WORK]} or @code{SET AUTOCOMMIT=0}, you must
use the MySQL binary log for backups instead of the old update log.

The binary logging is done immediately after a query completes but before
any locks are released or any commit is done. This ensures that the log
will be logged in the execution order.

All updates (@code{UPDATE}, @code{DELETE} or @code{INSERT}) that change
a transactional table (like BDB tables) are cached until a @code{COMMIT}.
Any updates to a non-transactional table are stored in the binary log at
once.  Every thread will, on start, allocate a buffer of
@code{binlog_cache_size} to buffer queries.  If a query is bigger than
this, the thread will open a temporary file to handle the bigger cache.
The temporary file will be deleted when the thread ends.

The @code{max_binlog_cache_size} can be used to restrict the total size used
to cache a multi-transaction query.

If you are using the update or binary log, concurrent inserts will
not work together with @code{CREATE ... INSERT} and @code{INSERT ... SELECT}.
This is to ensure that you can recreate an exact copy of your tables by
applying the log on a backup.


@node Slow query log, Log file maintenance, Binary log, Log Files
@subsection The Slow Query Log

@cindex slow query log
@cindex files, slow query log

When started with the @code{--log-slow-queries[=file_name]} option,
@code{mysqld} writes a log file containing all SQL commands that took
more than @code{long_query_time} to execute. The time to get the initial
table locks are not counted as execution time.

The slow query log is logged after the query is executed and after all
locks has been released. This may be different from the order in which
the statements are executed.

If no file name is given, it defaults to the name of the host machine
suffixed with @code{-slow.log}. If a filename is given, but doesn't
contain a path, the file is written in the data directory.

The slow query log can be used to find queries that take a long time to
execute and are thus candidates for optimisation. With a large log, that
can become a difficult task. You can pipe the slow query log through the
@code{mysqldumpslow} command to get a summary of the queries which
appear in the log.

You are using @code{--log-long-format} then also queries that are not
using indexes are printed. @xref{Command-line options}.


@node Log file maintenance,  , Slow query log, Log Files
@subsection Log File Maintenance

@cindex files, log
@cindex maintaining, log files
@cindex log files, maintaining

MySQL has a lot of log files which make it easy to see what is
going. @xref{Log Files}. One must however from time to time clean up
after @code{MysQL} to ensure that the logs don't take up too much disk
space.

When using MySQL with log files, you will, from time to time,
want to remove/backup old log files and tell MySQL to start
logging on new files.  @xref{Backup}.

On a Linux (@code{Redhat}) installation, you can use the
@code{mysql-log-rotate} script for this. If you installed MySQL
from an RPM distribution, the script should have been installed
automatically.  Note that you should be careful with this if you are using
the log for replication!

On other systems you must install a short script yourself that you
start from @code{cron} to handle log files.

You can force MySQL to start using new log files by using
@code{mysqladmin flush-logs} or by using the SQL command @code{FLUSH LOGS}.
If you are using MySQL Version 3.21 you must use @code{mysqladmin refresh}.

The above command does the following:

@itemize @bullet
@item
If standard logging (@code{--log}) or slow query logging
(@code{--log-slow-queries}) is used, closes and reopens the log file
(@file{mysql.log} and @file{`hostname`-slow.log} as default).
@item
If update logging (@code{--log-update}) is used, closes the update log and
opens a new log file with a higher sequence number.
@end itemize

If you are using only an update log, you only have to flush the logs and then
move away the old update log files to a backup.
If you are using the normal logging, you can do something like:

@example
shell> cd mysql-data-directory
shell> mv mysql.log mysql.old
shell> mysqladmin flush-logs
@end example

and then take a backup and remove @file{mysql.old}.


@node Replication,  , Log Files, MySQL Database Administration
@section Replication in MySQL

@cindex replication
@cindex increasing, speed
@cindex speed, increasing
@cindex databases, replicating

@menu
* Replication Intro::           Introduction
* Replication Implementation::  Replication Implementation Overview
* Replication HOWTO::           How To Set Up Replication
* Replication Features::        Replication Features and Known Problems
* Replication Options::         Replication Options in @file{my.cnf}
* Replication SQL::             SQL Commands Related to Replication
* Replication FAQ::             Replication FAQ
* Replication Problems::        Troubleshooting Replication
@end menu

This section describes the various replication features in MySQL.
It serves as a reference to the options available with replication.
You will be introduced to replication and learn how to implement it.
Toward the end, there are some frequently asked questions and descriptions
of problems and how to solve them.

We suggest that you visit our website at @uref{http://www.mysql.com/}
often and read updates to this section. Replication is constantly being
improved, and we update the manual frequently with the most current
information.

@node Replication Intro, Replication Implementation, Replication, Replication
@subsection Introduction

One way replication can be used is to increase both robustness and
speed. For robustness you can have two systems and can switch to the backup if
you have problems with the master. The extra speed is achieved by
sending a part of the non-updating queries to the replica server. Of
course this only works if non-updating queries dominate, but that is the
normal case.

Starting in Version 3.23.15, MySQL supports one-way replication
internally. One server acts as the master, while the other acts as the
slave. Note that one server could play the roles of master in one pair
and slave in the other. The master server keeps a binary log of updates
(@pxref{Binary log}) and an index file to binary logs to keep track of
log rotation.  The slave, upon connecting, informs the master where it
left off since the last successfully propagated update, catches up on
the updates, and then blocks and waits for the master to notify it of
the new updates.

Note that if you are replicating a database, all updates to this
database should be done through the master!

Another benefit of using replication is that one can get live backups of
the system by doing a backup on a slave instead of doing it on the
master. @xref{Backup}.


@node Replication Implementation, Replication HOWTO, Replication Intro, Replication
@subsection Replication Implementation Overview

@cindex master-slave setup

MySQL replication is based on the server keeping track of all
changes to your database (updates, deletes, etc) in the binary
log (@pxref{Binary log}) and the slave server(s) reading the saved
queries from the master server's binary log so that the slave can
execute the same queries on its copy of the data.

It is @strong{very important} to realise that the binary log is simply a
record starting from a fixed point in time (the moment you enable binary
logging). Any slaves which you set up will need copies of all the data
from your master as it existed the moment that you enabled binary
logging on the master. If you start your slaves with data that doesn't
agree with what was on the master @strong{when the binary log was
started}, your slaves may fail.

Starting in 4.0.0, one can use @code{LOAD DATA FROM MASTER} to set up
a slave. Note that 4.0.0 slaves cannot communicate with 3.23 masters, but 4.0.1
and later version slaves can. 3.23 slave cannot talk to 4.0 master.

You must also be aware that @code{LOAD DATA FROM MASTER} currently works only
if all the tables on the master are @code{MyISAM} type, and will acuire a
global read lock, so no writes are possible while the tables are being
transferred from the master. This limitation is of a temporary nature, and is
due to the fact that we have not yet implemented hot lock-free table backup.
It will be removed in the future 4.0 branch versions once we implemented hot
backup enabling @code{LOAD DATA FROM MASTER} to work without blocking master
updates.

Due to the above limitation, we recommend that at this point  you use
@code{LOAD DATA FROM MASTER} only if the dataset on the master is relatively
small, or if a prolonged read lock on the master is acceptable. While the
actual speed of @code{LOAD DATA FROM MASTER} may vary from system to system,
a good rule for a rough estimate of how long it is going to take is 1 second
per 1 MB of the datafile. You will get close to the estimate if both master
and slave are equivalent to 700 MHz Pentium, are connected through
100 MBit/s network, and your index file is about half the size of your data
file. Of course, your mileage will vary from system to system, the above rule
just gives you a rough order of magnitude estimate.

Once a slave is properly configured and running, it will simply connect
to the master and wait for updates to process. If the master goes away
or the slave loses connectivity with your master, it will keep trying to
connect every @code{master-connect-retry} seconds until it is able to
reconnect and resume listening for updates.

Each slave keeps track of where it left off. The master server has no
knowledge of how many slaves there are or which ones are up-to-date at
any given time.

The next section explains the master/slave setup process in more detail.


@node Replication HOWTO, Replication Features, Replication Implementation, Replication
@subsection How To Set Up Replication

Here is a quick description of how to set up complete replication on
your current MySQL server. It assumes you want to replicate all
your databases and have not configured replication before. You will need
to shutdown your master server briefly to complete the steps outlined
here.

While this method is the most straightforward way to set up a slave,
it is not the only one. For example, if you already have a snapshot
of the master, and
the master already has server id set and binary logging enabled, you can
set up a slave without shutting the master down or even blocking the updates.
For more details, please see @ref{Replication FAQ}.

If you want to become a real MySQL replication guru, we suggest that you
begin by studying, pondering, and trying all commands
mentioned in @ref{Replication SQL}. You should also familiarize yourself
with replication startup options in @file{my.cnf} in
@ref{Replication Options}.

@enumerate
@item
Make sure you have a recent version of MySQL installed on the master
and slave(s).

Use Version 3.23.29 or higher. Previous releases used a different binary
log format and had bugs which have been fixed in newer releases. Please,
do not report bugs until you have verified that the problem is present
in the latest release.

@item
Set up special a replication user on the master with the @code{FILE}
privilege and permission to connect from all the slaves. If the user is
only doing replication (which is recommended), you don't need to grant any
additional privileges.

For example, to create a user named @code{repl} which can access your
master from any host, you might use this command:

@example
mysql> GRANT FILE ON *.* TO repl@@"%" IDENTIFIED BY '<password>';
@end example

@item
Shut down MySQL on the master.

@example
mysqladmin -u root -p<password> shutdown
@end example

@item
Snapshot all the data on your master server.

The easiest way to do this (on Unix) is to simply use @strong{tar} to
produce an archive of your entire data directory. The exact data
directory location depends on your installation.

@example
tar -cvf /tmp/mysql-snapshot.tar /path/to/data-dir
@end example

Windows users can use @code{WinZIP} or similar software to create an
archive of the data directory.

@item
In @file{my.cnf} on the master add @code{log-bin} and
@code{server-id=unique number} to the @code{[mysqld]} section and
restart it. It is very important that the id of the slave is different from
the id of the master. Think of @code{server-id} as something similar
to the IP address - it uniquely identifies the server instance in the
community of replication partners.

@example
[mysqld]
log-bin
server-id=1
@end example

@item
Restart MySQL on the master.

@item
Add the following to @file{my.cnf} on the slave(s):

@example
master-host=<hostname of the master>
master-user=<replication user name>
master-password=<replication user password>
master-port=<TCP/IP port for master>
server-id=<some unique number between 2 and 2^32-1>
@end example

replacing the values in <> with what is relevant to your system.

@code{server-id} must be different for each server participating in
replication.  If you don't specify a server-id, it will be set to 1 if
you have not defined @code{master-host}, else it will be set to 2. Note
that in the case of @code{server-id} omission the master will refuse
connections from all slaves, and the slave will refuse to connect to a
master. Thus, omitting @code{server-id} is only good for backup with a
binary log.


@item
Copy the snapshot data into your data directory on your slave(s). Make
sure that the privileges on the files and directories are correct. The
user which MySQL runs as needs to be able to read and write to
them, just as on the master.

@item Restart the slave(s).

@end enumerate

After you have done the above, the slave(s) should connect to the master
and catch up on any updates which happened since the snapshot was taken.

If you have forgotten to set @code{server-id} for the slave you will get
the following error in the error log file:

@example
Warning: one should set server_id to a non-0 value if master_host is set.
The server will not act as a slave.
@end example

If you have forgotten to do this for the master, the slaves will not be
able to connect to the master.

If a slave is not able to replicate for any reason, you will find error
messages in the error log on the slave.

Once a slave is replicating, you will find a file called
@file{master.info} in the same directory as your error log. The
@file{master.info} file is used by the slave to keep track of how much
of the master's binary log it has processed. @strong{Do not} remove or
edit the file, unless you really know what you are doing. Even in that case,
it is preferred that you use @code{CHANGE MASTER TO} command.

@node Replication Features, Replication Options, Replication HOWTO, Replication
@subsection Replication Features and Known Problems

@cindex options, replication
@cindex @file{my.cnf} file
@cindex files,@file{my.cnf}

Here is an explanation of what is supported and what is not:

@itemize @bullet
@item
Replication will be done correctly with @code{AUTO_INCREMENT},
@code{LAST_INSERT_ID()}, and @code{TIMESTAMP} values.
@item
@code{RAND()} in updates does not replicate properly. Use
@code{RAND(some_non_rand_expr)} if you are replicating updates with
@code{RAND()}. You can, for example, use @code{UNIX_TIMESTAMP()} for the
argument to @code{RAND()}.
@item
You have to use the same character set (@code{--default-character-set})
on the master and the slave. If not, you may get duplicate key errors on
the slave, because a key that is regarded as unique on the master may
not be that in the other character set.
@item
In 3.23, @code{LOAD DATA INFILE} will be handled properly as long as the file
still resides on the master server at the time of update
propagation. @code{LOAD LOCAL DATA INFILE} will be skipped. In 4.0, this
limitation is not present - all forms of @code{LOAD DATA INFILE} are properly
replicated.
@item
Update queries that use user variables are not replication-safe (yet).
@item
@code{FLUSH} commands are not stored in the binary log and are because
of this not replicated to the slaves. This is not normally a problem as
@code{FLUSH} doesn't change anything. This does however mean that if you
update the MySQL privilege tables directly without using the
@code{GRANT} statement and you replicate the @code{mysql} privilege
database, you must do a @code{FLUSH PRIVILEGES} on your slaves to put
the new privileges into effect.
@item
Temporary tables starting in 3.23.29 are replicated properly with the
exception of the case when you shut down slave server ( not just slave thread),
you have some temporary tables open, and they are used in subsequent updates.
To deal with this problem shutting down the slave, do @code{SLAVE STOP},
check @code{Slave_open_temp_tables} variable to see if it is 0, then issue
@code{mysqladmin shutdown}. If the number is not 0, restart the slave thread
with @code{SLAVE START} and see
if you have better luck next time. There will be a cleaner solution, but it
has to wait until version 4.0.
In earlier versions temporary tables are not replicated properly - we
recommend that you either upgrade, or execute @code{SET SQL_LOG_BIN=0} on
your clients before all queries with temp tables.
@item
MySQL only supports one master and many slaves. In 4.x, we will
add a voting algorithm to automatically change master if something goes
wrong with the current master. We will also introduce 'agent' processes
to help do load balancing by sending select queries to different
slaves.
@item
Starting in Version 3.23.26, it is safe to connect servers in a circular
master-slave relationship with @code{log-slave-updates} enabled.
Note, however, that many queries will not work right in this kind of
setup unless your client code is written to take care of the potential
problems that can happen from updates that occur in different sequence
on different servers.

This means that you can do a setup like the following:

@example
A -> B -> C -> A
@end example

This setup will only works if you only do non conflicting updates
between the tables.  In other words, if you insert data in A and C, you
should never insert a row in A that may have a conflicting key with a
row insert in C.  You should also not update the sam rows on two servers
if the order in which the updates are applied matters.

Note that the log format has changed in Version 3.23.26 so that
pre-3.23.26 slaves will not be able to read it.
@item
If the query on the slave gets an error, the slave thread will
terminate, and a message will appear in the @file{.err} file. You should
then connect to the slave manually, fix the cause of the error (for
example, non-existent table), and then run the @code{SLAVE START} SQL
command (available starting in Version 3.23.16). In Version 3.23.15, you
will have to restart the server.
@item
If connection to the master is lost, the slave will retry immediately,
and then in case of failure every @code{master-connect-retry} (default
60) seconds. Because of this, it is safe to shut down the master, and
then restart it after a while. The slave will also be able to deal with
network connectivity outages.
@item
Shutting down the slave (cleanly) is also safe, as it keeps track of
where it left off.  Unclean shutdowns might produce problems, especially
if disk cache was not synced before the system died. Your system fault
tolerance will be greatly increased if you have a good UPS.
@item
If the master is listening on a non-standard port, you will also need to
specify this with @code{master-port} parameter in @file{my.cnf} .
@item
In Version 3.23.15, all of the tables and databases will be
replicated. Starting in Version 3.23.16, you can restrict replication to
a set of databases with @code{replicate-do-db} directives in
@file{my.cnf} or just exclude a set of databases with
@code{replicate-ignore-db}. Note that up until Version 3.23.23, there was a bug
that did not properly deal with @code{LOAD DATA INFILE} if you did it in
a database that was excluded from replication.
@item
Starting in Version 3.23.16, @code{SET SQL_LOG_BIN = 0} will turn off
replication (binary) logging on the master, and @code{SET SQL_LOG_BIN =
1} will turn it back on -- you must have the @strong{process} privilege to do
this.
@item
Starting in Version 3.23.19, you can clean up stale replication leftovers when
something goes wrong and you want a clean start with @code{FLUSH MASTER}
and @code{FLUSH SLAVE} commands. In Version 3.23.26 we have renamed them to
@code{RESET MASTER} and @code{RESET SLAVE} respectively to clarify
what they do. The old @code{FLUSH} variants still work, though, for
compatibility.

@item
Starting in Version 3.23.23, you can change masters and adjust log position
with @code{CHANGE MASTER TO}.
@item
Starting in Version 3.23.23, you tell the master that updates in certain
databases should not be logged to the binary log with @code{binlog-ignore-db}.
@item
Starting in Version 3.23.26, you can use @code{replicate-rewrite-db} to tell
the slave to apply updates from one database on the master to the one
with a different name on the slave.
@item
Starting in Version 3.23.28, you can use @code{PURGE MASTER LOGS TO 'log-name'}
to get rid of old logs while the slave is running.
@item
Due to the non-transactional nature of MyISAM tables, it is possible to have
a query that will only partially update a table and return an error code. This
can happen, for example, on a multi-row insert that has one row violating a
key constraint, or if a long update query is killed after updating some of the
rows. If that happens on the master, the slave thread will exit and wait for
the DBA to decide what to do about it unless the error code is legitimate and
the query execution results in the same error code. If this error code
validation behaviour is not desirable, some ( or all) errors could be masked
out with @code{slave-skip-errors} option starting in Version 3.23.47.
@item
While individual tables can be excluded from replication with
@code{replicate-do-table}/@code{replicate-ignore-table} or
@code{replicate-wild-do-table}/@code{replicate-wild-ignore-table}, there
are currently some design deficiencies that in some rather rare cases
produce unexpected results. The replication protocol does not inform the
slave explicitly which tables are going to be modified by the query -- so
the slave has to parse the query to know this. To avoid redundant
parsing for queries that will end up actually being executed, table
exclusion is currently implemented by sending the query to the standard
MySQL parser, which will short-circuit the query and report success if
it detects that the table should be ignored. In addition to several
inefficiencies, this approach is also more bug prone, and there are two
known bugs as of Version 3.23.49 -- because the parser automatically opens
the table when parsing some queries the ignored table has to exist on
the slave. The other bug is that if the ignored table gets partially
updated, the slave thread will not notice that the table actually should
have been ignored and will suspend the replication process. While the
above bugs are conceptually very simple to fix, we have not yet found a way
to do this without a sigficant code change that would compromize the stability
status of 3.23 branch. There exists a workaround for both if in the rare case
it happens to affect your application -- use @code{slave-skip-errors}.

@end itemize


@node Replication Options, Replication SQL, Replication Features, Replication
@subsection Replication Options in @file{my.cnf}

If you are using replication, we recommend that you use MySQL Version
3.23.30 or later. Older versions work, but they do have some bugs and are
missing some features. Some of the options mentioned here may not be available in
your version if it is not the most recent one. For all options specific to
the 4.0 branch, there is a note indicating so. Otherwise, if you discover
that the option you are interested in is not available in your 3.23 version,
and you really need it, please upgrade to the most recent 3.23 branch.

Please be aware that 4.0 branch is still in alpha, so some things may not be
working as smoothly as you would like. If you really would like to try
the new features of 4.0, we recommend you do it in such a way that in
case there is a problem  your mission critical applications will not be
disrupted.

On both master and slave you need to use the @code{server-id} option.
This sets an unique replication id. You should pick a unique value in the
range between 1 to 2^32-1 for each master and slave.
Example: @code{server-id=3}

The following table describes the options you can use for the @code{MASTER}:

@multitable @columnfractions .38 .62
@item @strong{Option} @tab @strong{Description}

@item @code{log-bin=filename} @tab
Write to a binary update log to the specified location. Note that if you
give it a parameter with an extension (for example,
@code{log-bin=/mysql/logs/replication.log} ) versions up to 3.23.24 will
not work right during replication if you do @code{FLUSH LOGS} . The
problem is fixed in Version 3.23.25. If you are using this kind of log
name, @code{FLUSH LOGS} will be ignored on binlog. To clear the log, run
@code{FLUSH MASTER}, and do not forget to run @code{FLUSH SLAVE} on all
slaves. In Versions 3.23.26 and later, you should use
@code{RESET MASTER} and @code{RESET SLAVE}

@item @code{log-bin-index=filename} @tab
Because the user could issue the @code{FLUSH LOGS} command, we need to
know which log is currently active and which ones have been rotated out
and in what sequence. This information is stored in the binary log index file.
The default is @file{`hostname`.index}. You should not need to change this.

Example: @code{log-bin-index=db.index}

@item @code{sql-bin-update-same} @tab
If set, setting @code{SQL_LOG_BIN} to a value will automatically set
@code{SQL_LOG_UPDATE} to the same value and vice versa.

@item @code{binlog-do-db=database_name} @tab
Tells the master that it should log updates to the binary log if the
current database is @code{database_name}. All other databases are ignored.
Note that if you use this, you should ensure that you do updates only in
the current database.

Example: @code{binlog-do-db=sales}

@item @code{binlog-ignore-db=database_name} @tab
Tells the master that updates where the current database is
@code{database_name} should not be stored in the binary log.  Note that if
you use this, you should ensure that you do updates only in the current
database.

Example: @code{binlog-ignore-db=accounting}
@end multitable

The following table describes the options you can use for the @code{SLAVE}:

@multitable @columnfractions .38 .62
@item @strong{Option} @tab @strong{Description}

@item @code{master-host=host} @tab
Master hostname or IP address for replication. If not set, the slave
thread will not be started. Note that the setting of @code{master-host}
will be ignored if there exists a valid @file{master.info} file. Probably a
better name for this options would have been something like
@code{bootstrap-master-host}, but it is too late to change now.

Example: @code{master-host=db-master.mycompany.com}

@item @code{master-user=username} @tab
The username the slave thread will use for authentication when connecting to
the master. The user must have the @strong{file} privilege. If the master user
is not set, user @code{test} is assumed. The value in @file{master.info} will
take precedence if it can be read.

Example: @code{master-user=scott}

@item @code{master-password=password} @tab
The password the slave thread will authenticate with when connecting to
the master. If not set, an empty password is assumed.The value in
@file{master.info} will take precedence if it can be read.

Example: @code{master-password=tiger}

@item @code{master-port=portnumber} @tab
The port the master is listening on. If not set, the compiled setting of
@code{MYSQL_PORT} is assumed. If you have not tinkered with
@code{configure} options, this should be 3306. The value in
@file{master.info} will take precedence if it can be read.

Example: @code{master-port=3306}

@item @code{master-connect-retry=seconds} @tab
The number of seconds the slave thread will sleep before retrying to
connect to the master in case the master goes down or the connection is
lost.  Default is 60.

Example: @code{master-connect-retry=60}

@item @code{master-ssl} @tab
Available after 4.0.0. Turn SSL on for replication. Be warned that is
this is a relatively new feature.

Example: @code{master-ssl}

@item @code{master-ssl-key} @tab
Available after 4.0.0. Master SSL keyfile name. Only applies if you have
enabled @code{master-ssl}.

Example: @code{master-ssl-key=SSL/master-key.pem}

@item @code{master-ssl-cert} @tab
Available after 4.0.0. Master SSL certificate file name. Only applies if
you have enabled @code{master-ssl}.


Example: @code{master-ssl-key=SSL/master-cert.pem}

@item @code{master-info-file=filename} @tab
The location of the file that remembers where we left off on the master
during the replication process. The default is @file{master.info} in the data
directory. You should not need to change this.

Example: @code{master-info-file=master.info}

@item @code{report-host} @tab
Available after 4.0.0. Hostname or IP of the slave to be reported to to
the master during slave registration. Will appear in the output of
@code{SHOW SLAVE HOSTS}. Leave unset if you do not want the slave to
register itself with the master. Note that it is not sufficient for the
master to simply read the IP of the slave off the socket once the slave
connects. Due to @code{NAT} and other routing issues, that IP may not be
valid for connecting to the slave from the master or other hosts.

Example: @code{report-host=slave1.mycompany.com}

@item @code{report-port} @tab
Available after 4.0.0. Port for connecting to slave reported to the
master during slave registration.  Set it only if the slave is listening
on a non-default port or if you have a special tunnel from the master or
other clients to the slave. If not sure, leave this option unset.

@item @code{replicate-do-table=db_name.table_name} @tab
Tells the slave thread to restrict replication to the specified table.
To specify more than one table, use the directive multiple times, once
for each table.  This will work for cross-database updates, in
contrast to @code{replicate-do-db}.

Example: @code{replicate-do-table=some_db.some_table}

@item @code{replicate-ignore-table=db_name.table_name} @tab
Tells the slave thread to not replicate to the specified table. To
specify more than one table to ignore, use the directive multiple times,
once for each table. This will work for cross-datbase updates,
in contrast to @code{replicate-ignore-db}.

Example: @code{replicate-ignore-table=db_name.some_table}

@item @code{replicate-wild-do-table=db_name.table_name} @tab
Tells the slave thread to restrict replication to the tables that match
the specified wildcard pattern.  To specify more than one table, use the
directive multiple times, once for each table.  This will work for
cross-database updates.

Example: @code{replicate-wild-do-table=foo%.bar%} will replicate only updates
to tables in all databases that start with @code{foo} and whose table names
start with @code{bar}.

@item @code{replicate-wild-ignore-table=db_name.table_name} @tab
Tells the slave thread to not replicate to the tables that match the
given wildcard pattern. To specify more than one table to ignore, use
the directive multiple times, once for each table. This will work for
cross-database updates.

Example: @code{replicate-wild-ignore-table=foo%.bar%} will not do updates
to tables in databases that start with @code{foo} and whose table names start
with @code{bar}.

@item @code{replicate-ignore-db=database_name} @tab
Tells the slave thread to not replicate to the specified database. To
specify more than one database to ignore, use the directive multiple
times, once for each database. This option will not work if you use cross
database updates. If you need cross database updates to work, make sure
you have 3.23.28 or later, and use
@code{replicate-wild-ignore-table=db_name.%}.

Example: @code{replicate-ignore-db=some_db}

@item @code{replicate-do-db=database_name} @tab

Tells the slave thread to restrict replication to the specified
database.  To specify more than one database, use the directive multiple
times, once for each database. Note that this will only work if you do
not use cross-database queries such as @code{UPDATE some_db.some_table
SET foo='bar'} while having selected a different or no database. If you
need cross database updates to work, make sure you have 3.23.28 or
later, and use @code{replicate-wild-do-table=db_name.%}.

Example: @code{replicate-do-db=some_db}

@item @code{log-slave-updates} @tab
Tells the slave to log the updates from the slave thread to the binary
log. Off by default. You will need to turn it on if you plan to
daisy-chain the slaves.

@item @code{replicate-rewrite-db=from_name->to_name} @tab
Updates to a database with a different name than the original.

Example: @code{replicate-rewrite-db=master_db_name->slave_db_name}

@item @code{slave-skip-errors= [err_code1,err_code2,... | all]} @tab

Available only in 3.23.47 and later. Tells the slave thread to continue
replication when a query returns an error from the provided
list. Normally, replication will discontinue when an error is
encountered, giving the user a chance to resolve the inconsistency in the
data manually. Do not use this option unless you fully understand why
you are getting the errors.  If there are no bugs in your
replication setup and client programs, and no bugs in MySQL itself, you
should never get an abort with error. Indiscriminate use of this option
will result in slaves being hopelessly out of sync with the master and
you having no idea how the problem happened.

For error codes, you should use the numbers provided by the error message in
your slave error log and in the output of @code{SHOW SLAVE STATUS}. Full list
of error messages can be found in the source distribution in
@file{Docs/mysqld_error.txt}.

You can (but should not) also use a very non-recommended value of @code{all}
which will ignore all error messages and keep barging along regardless.
Needless to say, if you use it, we make no promises regarding your data
integrity. Please do not complain if your data on the slave is not anywhere
close to what it is on the master in this case -- you have been warned.

Example:

@code{slave-skip-errors=1062,1053} or @code{slave-skip-errors=all}

@item @code{skip-slave-start} @tab
Tells the slave server not to start the slave on the startup.  The user
can start it later with @code{SLAVE START}.

@item @code{slave_net_timeout=#} @tab
Number of seconds to wait for more data from the master before aborting
the read.
@end multitable


@node Replication SQL, Replication FAQ, Replication Options, Replication
@subsection SQL Commands Related to Replication

@cindex SQL commands, replication
@cindex commands, replication
@cindex replication, commands

Replication can be controlled through the SQL interface. Here is the
summary of commands:

@multitable @columnfractions .45 .55
@item @strong{Command} @tab @strong{Description}

@item @code{SLAVE START}
 @tab Starts the slave thread. (Slave)

@item @code{SLAVE STOP}
 @tab Stops the slave thread. (Slave)

@item @code{SET SQL_LOG_BIN=0}
 @tab Disables update logging if the user has the @strong{process} privilege.
 Ignored otherwise. (Master)

@item @code{SET SQL_LOG_BIN=1}
 @tab Re-enables update logging if the user has the @strong{process} privilege.
 Ignored otherwise. (Master)

@item @code{SET SQL_SLAVE_SKIP_COUNTER=n}
 @tab Skip the next @code{n} events from the master. Only valid when
the slave thread is not running, otherwise, gives an error. Useful for
recovering from replication glitches.

@item @code{RESET MASTER}
 @tab Deletes all binary logs listed in the index file, resetting the binlog
index file to be empty. In pre-3.23.26 versions, use @code{FLUSH MASTER}
(Master)

@item @code{RESET SLAVE}
 @tab Makes the slave forget its replication position in the master
logs. In pre 3.23.26 versions the command was called
@code{FLUSH SLAVE}(Slave)

@item @code{LOAD TABLE tblname FROM MASTER}
 @tab Downloads a copy of the table from master to the slave. Implemented
 mainly for debugging of @code{LOAD DATA FROM MASTER}, but some ``gourmet''
 users might find it useful for other things. Do not use it if you consider
 yourself the average ``non-hacker'' type user. (Slave)

@item @code{LOAD DATA FROM MASTER} @tab
Available starting in 4.0.0. Takes a snapshot of the master and copies
it to the slave. Updates the values of @code{MASTER_LOG_FILE} and
@code{MASTER_LOG_POS} so that the slave will start replicating from the
correct position. Will honor table and database exclusion rules
specified with @code{replicate-*} options. So far works only with
@code{MyISAM} tables and acquires a global read lock on the master while
taking the snapshot. In the future it is planned to make it work with
@code{InnoDB} tables and to remove the need for global read lock using
the non-blocking online backup feature.

@item @code{CHANGE MASTER TO master_def_list}
 @tab Changes the master parameters to the values specified in
@code{master_def_list} and restarts the slave thread. @code{master_def_list}
is a comma-separated list of @code{master_def} where @code{master_def} is
one of the following: @code{MASTER_HOST}, @code{MASTER_USER},
@code{MASTER_PASSWORD}, @code{MASTER_PORT}, @code{MASTER_CONNECT_RETRY},
@code{MASTER_LOG_FILE}, @code{MASTER_LOG_POS}. For example:

@example
CHANGE MASTER TO
  MASTER_HOST='master2.mycompany.com',
  MASTER_USER='replication',
  MASTER_PASSWORD='bigs3cret',
  MASTER_PORT=3306,
  MASTER_LOG_FILE='master2-bin.001',
  MASTER_LOG_POS=4;
@end example

You only need to specify the values that need to be changed. The values that
you omit will stay the same with the exception of when you change the host or
the port. In that case, the slave will assume that since you are connecting to
a different host or a different port, the master is different. Therefore, the
old values of log and position are not applicable anymore, and will
automatically be reset to an empty string and 0, respectively (the start
values). Note that if you restart the slave, it will remember its last master.
If this is not desirable, you should delete the @file{master.info} file before
restarting, and the slave will read its master from @file{my.cnf} or the
command-line.

This command is useful for setting up a slave when you have the snapshot of
the master and have recorded the log and the offset on the master that the
snapshot corresponds to. You can run
 @code{CHANGE MASTER TO MASTER_LOG_FILE='log_name_on_master',
 MASTER_LOG_POS=log_offset_on_master} on the slave after restoring the
snapshot.

(Slave)

@item @code{SHOW MASTER STATUS} @tab Provides status information on the binlog of the master. (Master)

@item @code{SHOW SLAVE HOSTS} @tab Available after 4.0.0. Gives a
listing of slaves currently registered with the master (Master)

@item @code{SHOW SLAVE STATUS} @tab Provides status information on essential parameters of the slave thread. (Slave)

@item @code{SHOW MASTER LOGS} @tab Only available starting in Version
3.23.28. Lists the binary logs on the master. You should use this
command prior to @code{PURGE MASTER LOGS TO} to find out how far you
should go. (Master)

@item @code{SHOW BINLOG EVENTS [ IN 'logname' ] [ FROM pos ]
 [ LIMIT [offset,] rows ] } @tab
Shows the events in the binary update log. Primarily used for
testing/debugging, but can also be used by regular clients that for some
reason need to read the binary log contents. (Master)

@item @code{SHOW NEW MASTER FOR SLAVE WITH MASTER_LOG_FILE='logfile' AND
 MASTER_LOG_POS=pos AND
MASTER_LOG_SEQ=log_seq AND MASTER_SERVER_ID=server_id} @tab
This command is used when a slave of a possibly dead/unavailable master
needs to be switched to replicate off another slave that has been
replicating the same master. The command will return recalculated
replication coordinates, and the output can be used in a subsequent
@code{CHANGE MASTER TO} command. Normal users should never need to run
this command. It is primarily reserved for internal use by the fail-safe
replication code. We may later change the syntax if we find a more
intuitive way to describe this operation.

@item @code{PURGE MASTER LOGS TO 'logname'}
 @tab  Available starting in Version 3.23.28. Deletes all the
replication logs that are listed in the log
index as being prior to the specified log, and removes them from the
log index, so that the given log now becomes the first. Example:

@example
PURGE MASTER LOGS TO 'mysql-bin.010'
@end example

This command will do nothing and fail with an error if you have an
active slave that is currently reading one of the logs you are trying to
delete. However, if you have a dormant slave, and happen to purge one of
the logs it wants to read, the slave will be unable to replicate once it
comes up.  The command is safe to run while slaves are replicating -- you
do not need to stop them.

You must first check all the slaves with @code{SHOW SLAVE STATUS} to
see which log they are on, then do a listing of the logs on the
master with @code{SHOW MASTER LOGS}, find the earliest log among all
the slaves (if all the slaves are up to date, this will be the
last log on the list), backup all the logs you are about to delete
(optional) and purge up to the target log.

@end multitable


@node Replication FAQ, Replication Problems, Replication SQL, Replication
@subsection Replication FAQ

@strong{Q}: How do I configure a slave if the master is already running
and I do not want to stop it?

@strong{A}: There are several options. If you have taken a backup of the
master at some point and recorded the binlog name and offset ( from the
output of @code{SHOW MASTER STATUS} ) corresponding to the snapshot, do
the following:

@itemize @bullet
@item
Make sure unique server id is assigned to the slave.
@item
Execute @code{CHANGE MASTER TO MASTER_HOST='master-host-name',
 MASTER_USER='master-user-name', MASTER_PASSWORD='master-pass',
 MASTER_LOG_FILE='recorded-log-name', MASTER_LOG_POS=recorded_log_pos}
@item
Execute @code{SLAVE START}
@end itemize

If you do not have a backup of the master already, here is a quick way to
do it consistently:

@itemize @bullet
@item
@code{FLUSH TABLES WITH READ LOCK}
@item
@code{gtar zcf /tmp/backup.tar.gz /var/lib/mysql} ( or a variation of this)
@item
@code{SHOW MASTER STATUS} - make sure to record the output - you will need it
later
@item
@code{UNLOCK TABLES}
@end itemize

Afterwards, follow the instructions for the case when you have a snapshot and
have recorded the log name and offset. You can use the same snapshot to set up
several slaves. As long as the binary logs of the master are left intact, you
can wait as long as several days or in some cases maybe a month to set up a
slave once you have the snapshot of the master. In theory the waiting gap can
be infinite. The two practical limitations is the diskspace of the master
getting filled with old logs, and the amount of time it will take the slave to
catch up.

In version 4.0.0 and newer, you can also use @code{LOAD DATA FROM
MASTER}.  This is a convenient command that will take a snapshot,
restore it to the slave, and adjust the log name and offset on the slave
all at once. In the future, @code{LOAD DATA FROM MASTER} will be the
recommended way to set up a slave.  Be warned, howerver, that the read
lock may be held for a long time if you use this command. It is not yet
implemented as efficiently as we would like to have it. If you have
large tables, the preferred method at this time is still with a local
@code{tar} snapshot after executing @code{FLUSH TABLES WITH READ LOCK}.

@strong{Q}: Does the slave need to be connected to the master all the time?

@strong{A}: No, it does not. You can have the slave go down or stay
disconnected for hours or even days, then reconnect, catch up on the
updates, and then disconnect or go down for a while again. So you can,
for example, use master-slave setup over a dial-up link that is up only
for short periods of time. The implications of that are that at any
given time the slave is not guaranteed to be in sync with the master
unless you take some special measures. In the future, we will have the
option to block the master until at least one slave is in sync.

@strong{Q}: How do I force the master to block updates until the slave catches
up?

@strong{A}: Execute the following commands:

@itemize @bullet
@item
Master: @code{FLUSH TABLES WITH READ LOCK}
@item
Master: @code{SHOW MASTER STATUS} - record the log name and the offset
@item
Slave: @code{SELECT MASTER_POS_WAIT('recorded_log_name', recorded_log_offset)}
When the select returns, the slave is currently in sync with the master
@item
Master: @code{UNLOCK TABLES} - now the master will continue updates.
@end itemize

@cindex @code{Binlog_Dump}
@strong{Q}: Why do I sometimes see more than one @code{Binlog_Dump} thread on
the master after I have restarted the slave?

@strong{A}: @code{Binlog_Dump} is a continuous process that is handled by the
server in the following way:

@itemize @bullet
@item
Catch up on the updates.
@item
Once there are no more updates left, go into @code{pthread_cond_wait()},
from which we can be awakened either by an update or a kill.
@item
On wake up, check the reason. If we are not supposed to die, continue
the @code{Binlog_dump} loop.
@item
If there is some fatal error, such as detecting a dead client,
terminate the loop.
@end itemize

So if the slave thread stops on the slave, the corresponding
@code{Binlog_Dump} thread on the master will not notice it until after
at least one update to the master (or a kill), which is needed to wake
it up from @code{pthread_cond_wait()}.  In the meantime, the slave
could have opened another connection, which resulted in another
@code{Binlog_Dump} thread.

The above problem should not be present in Version 3.23.26 and later
versions.  In Version 3.23.26 we added @code{server-id} to each
replication server, and now all the old zombie threads are killed on the
master when a new replication thread connects from the same slave

@strong{Q}: How do I rotate replication logs?

@strong{A}: In Version 3.23.28 you should use @code{PURGE MASTER LOGS
TO} command after determining which logs can be deleted, and optionally
backing them up first. In earlier versions the process is much more
painful, and cannot be safely done without stopping all the slaves in
the case that you plan to re-use log names. You will need to stop the
slave threads, edit the binary log index file, delete all the old logs,
restart the master, start slave threads, and then remove the old log files.


@strong{Q}: How do I upgrade on a hot replication setup?

@strong{A}: If you are upgrading  pre-3.23.26 versions, you should just
lock the master tables, let the slave catch up, then run @code{FLUSH
MASTER} on the master, and @code{FLUSH SLAVE} on the slave to reset the
logs, then restart new versions of the master and the slave. Note that
the slave can stay down for some time -- since the master is logging
all the updates, the slave will be able to catch up once it is up and
can connect.

After 3.23.26, we have locked the replication protocol for modifications, so
you can upgrade masters and slave on the fly to a newer 3.23 version and you
can have different versions of MySQL running on the slave and the
master, as long as they are both newer than 3.23.26.

@cindex replication, two-way
@strong{Q}: What issues should I be aware of when setting up two-way
replication?

@strong{A}: MySQL replication currently does not support any
locking protocol between master and slave to guarantee the atomicity of
a distributed (cross-server) update. In in other words, it is possible
for client A to make an update to  co-master 1, and in the meantime,
before it propagates to co-master 2, client B could make an update to
co-master 2 that will make the update of client A work differently than
it did on co-master 1. Thus when the update of client A will make it
to co-master 2, it will produce  tables that will be different from
what you have on co-master 1, even after all the updates from co-master
2 have also propagated. So you should not co-chain two servers in a
two-way replication relationship, unless you are sure that you updates
can safely happen in any order, or unless you take care of mis-ordered
updates somehow in the client code.


You must also realise that two-way replication actually does not improve
performance very much, if at all, as far as updates are concerned. Both
servers need to do the same amount of updates each, as you would have
one server do. The only difference is that there will be a little less
lock contention, because the updates originating on another server will
be serialised in one slave thread. This benefit, though, might be
offset by network delays.

@cindex performance, improving
@cindex increasing, performance
@strong{Q}: How can I use replication to improve performance of my system?

@strong{A}: You should set up one server as the master, and direct all
writes to it, and configure as many slaves as you have the money and
rackspace for, distributing the reads among the master and the slaves.
You can also start the slaves with @code{--skip-bdb},
@code{--low-priority-updates} and @code{--delay-key-write-for-all-tables}
to get speed improvements for the slave.  In this case the slave will
use non-transactional @code{MyISAM} tables instead of @code{BDB} tables
to get more speed.

@strong{Q}: What should I do to prepare my client code to use
performance-enhancing replication?

@strong{A}:
If the part of your code that is responsible for database access has
been properly abstracted/modularised, converting it to run with the
replicated setup should be very smooth and easy -- just change the
implementation of your database access to read from some slave or the
master, and to always write to the master. If your code does not have
this level of abstraction,
setting up a replicated system will give you an opportunity/motivation
to it clean up.
 You should start by creating a wrapper library
/module with the following functions:

@itemize @bullet
@item
@code{safe_writer_connect()}
@item
@code{safe_reader_connect()}
@item
@code{safe_reader_query()}
@item
@code{safe_writer_query()}
@end itemize

@code{safe_} means that the function will take care of handling all
the error conditions.

You should then convert your client code to use the wrapper library.
It may be a painful and scary process at first, but it will pay off in
the long run. All applications that follow the above pattern will be
able to take advantage of one-master/many slaves solution.  The
code will be a lot easier to maintain, and adding troubleshooting
options will be trivial. You will just need to modify one or two
functions, for example, to log how long each query took, or which
query, among your many thousands, gave you an error. If you have
written a lot of code already, you may want to automate the conversion
task by using Monty's @code{replace} utility, which comes with the
standard distribution of MySQL, or just write your own Perl script.
Hopefully, your code follows some recognisable pattern. If not, then
you are probably better off rewriting it anyway, or at least going
through and manually beating it into a pattern.

Note that, of course, you can use different names for the
functions. What is important is having unified interface for connecting
for reads, connecting for writes, doing a read, and doing a write.


@strong{Q}: When and how much can MySQL replication improve the performance
of my system?

@strong{A}: MySQL replication is most beneficial for a system
with frequent reads and not so frequent writes. In theory, by using a
one master/many slaves setup you can scale by adding more slaves until
you either run out of network bandwidth, or your update
load grows to the point
that the master cannot handle it.

In order to determine how many slaves you can get before the added
benefits begin to level out, and how much you can improve performance
of your site, you need to know your query patterns, and empirically
 (by benchmarking) determine the relationship between the throughput
on reads (reads per second, or @code{max_reads}) and on writes
@code{max_writes}) on a typical master and a typical slave. The
example here will show you a rather simplified calculation of what you
can get with replication for our imagined system.

Let's say our system load consists of 10% writes and 90% reads, and we
have determined that @code{max_reads} = 1200 - 2 * @code{max_writes},
or in other words, our system can do 1200 reads per second with no
writes, our average write is twice as slow as average read,
and the relationship is
linear. Let us suppose that our master and slave are of the same
capacity, and we have N slaves and 1 master. Then we have for each
server (master or slave):

@code{reads = 1200 - 2 * writes} (from bencmarks)

@code{reads = 9* writes / (N + 1) } (reads split, but writes go
to all servers)

@code{9*writes/(N+1) + 2 * writes = 1200}

@code{writes = 1200/(2 + 9/(N+1)}

So if N = 0, which means we have no replication, our system can handle
1200/11, about 109 writes per second (which means we will have 9 times
as many reads due to the nature of our application).

If N = 1, we can get up to 184 writes per second.

If N = 8, we get up to 400.

If N = 17, 480 writes.

Eventually as N approaches infinity (and our budget negative infinity),
we can get very close to 600 writes per second, increasing system
throughput about 5.5 times. However, with only 8 servers, we increased
it almost 4 times already.

Note that our computations assumed infinite network bandwidth, and
neglected several other factors that could turn out to be significant on
your system. In many cases, you may not be able to make a computation
similar to the one above that will accurately predict what will happen
on your system if you add N replication slaves. However, answering the
following questions should help you decided whether and how much, if at
all, the replication will improve the performance of your system:

@itemize @bullet
@item
What is the read/write ratio on your system?
@item
How much more write load can one server handle if you reduce the reads?
@item
How many slaves do you have bandwidth for on your network?
@end itemize

@strong{Q}: How can I use replication to provide redundancy/high
availability?

@strong{A}: With the currently available features, you would have to
set up a master and a slave (or several slaves), and write a script
that will monitor the
master to see if it is up, and instruct your applications and
the slaves of the master change in case of failure. Some suggestions:

@itemize @bullet
@item
To tell a slave to change the master use the @code{CHANGE MASTER TO} command.
@item
A good way to keep your applications informed as to the location of the
master is by having a dynamic DNS entry for the master.
With @code{bind} you can use @file{nsupdate} to dynamically update your DNS.
@item
You should run your slaves with the @code{log-bin} option and without
@code{log-slave-updates}. This way the slave will be ready to become a
master as soon as you issue @code{STOP SLAVE}; @code{RESET MASTER}, and
@code{CHANGE MASTER TO} on the other slaves. It will also help you catch
spurious updates that may happen because of misconfiguration of the
slave (ideally, you want to configure access rights so that no client
can update the slave, except for the slave thread) combined with the
bugs in your client programs (they should never update the slave
directly).

@end itemize

We are currently working on integrating an automatic master election
system into MySQL, but until it is ready, you will have to
create your own monitoring tools.


@node Replication Problems,  , Replication FAQ, Replication
@subsection Troubleshooting Replication

If you have followed the instructions, and your replication setup is not
working, first eliminate the user error factor by checking the following:

@itemize @bullet
@item
Is the master logging to the binary log? Check with @code{SHOW MASTER STATUS}.
If it is, @code{Position} will be non-zero. If not, verify that you have
given the master @code{log-bin} option and have set @code{server-id}.
@item
Is the slave running? Check with @code{SHOW SLAVE STATUS}. The answer is found
in @code{Slave_running} column. If not, verify slave options and check the
error log for messages.
@item
If the slave is running, did it establish connection with the master? Do
@code{SHOW PROCESSLIST}, find the thread with @code{system user} value in
@code{User} column and @code{none} in the @code{Host} column, and check the
@code{State} column. If it says @code{connecting to master}, verify the
privileges for the replication user on the master, master host name, your
DNS setup, whether the master is actually running, whether it is reachable
from the slave, and if all that seems okay, read the error logs.
@item
If the slave was running, but then stopped, look at SHOW SLAVE STATUS
output and check the error logs. It usually
happens when some query that succeeded on the master fails on the slave. This
should never happen if you have taken a proper snapshot of the master, and
never modify the data on the slave outside of the slave thread. If it does,
it is a bug, read below on how to report it.
@item
If a query on that succeeded on the master refuses to run on the slave, and
a full database resync ( the proper thing to do ) does not seem feasible,
try the following:
@itemize @minus
@item
First see if there is some stray record in the way. Understand how it got
there, then delete it and run @code{SLAVE START}
@item
If the above does not work or does not apply, try to understand if it would
be safe to make the update manually ( if needed) and then ignore the next
query from the master.
@item
If you have decided you can skip the next query, do
@code{SET SQL_SLAVE_SKIP_COUNTER=1; SLAVE START;} to skip a query that
does not use @code{AUTO_INCREMENT} or @code{LAST_INSERT_ID()},  or
@code{SET SQL_SLAVE_SKIP_COUNTER=2; SLAVE START;} otherwise. The reason
queries that use @code{AUTO_INCREMENT} or @code{LAST_INSERT_ID()}
are different is that they take two events in the binary log of the master.

@item
If you are sure the slave started out perfectly in sync with the master,
and no one has updated  the tables involved outside of slave thread,
report the bug, so
you will not have to do the above tricks again.
@end itemize
@item
Make sure you are not running into an old bug by upgrading to the most recent
version.
@item
If all else fails, read the error logs. If they are big,
@code{grep -i slave /path/to/your-log.err} on the slave. There is no
generic pattern to search for on the master, as the only errors it logs
are general system errors -- if it can, it will send the error to the slave
when things go wrong.
@end itemize

When you have determined that there is no user error involved, and replication
still either does not work at all or is unstable, it is time to start working
on a bug report. We need to get as much info as possible from you to be able
to track down the bug. Please do spend some time and effort preparing a good
bug report. Ideally, we would like to have a test case in the format found in
@code{mysql-test/t/rpl*} directory of the source tree. If you submit a test
case like that, you can expect a patch within a day or two in most cases,
although, of course, you mileage may vary depending on a number of factors.

The second best option is to write a simple program with easily configurable
connection arguments for the master and the slave that will demonstrate
the problem on our systems. You can write one in Perl or in C, depending
on which language you know better.

If you have one of the above ways to demonstrate the bug, use
@code{mysqlbug} to prepare a bug report and send it to
@email{bugs@@lists.mysql.com}. If you have a phantom -- a problem that
does occur but you cannot duplicate "at will":

@itemize @bullet
@item
Verify that there is no user error involved. For example, if you update the
slave outside of the slave thread, the data will be out of sync, and you can
have unique key violations on updates, in which case the slave thread will
stop and wait for you to clean up the tables manually to bring them in sync.
@item
Run slave with @code{log-slave-updates} and @code{log-bin} -- this will keep
a log of all updates on the slave.
@item
Save all evidence before resetting the replication. If we have no or only
sketchy information, it would take us a while to track down the problem. The
evidence you should collect is:
@itemize @minus
@item
All binary logs on the master
@item
All binary log on the slave
@item
The output of @code{SHOW MASTER STATUS} on the master at the time
you have discovered the problem
@item
The output of @code{SHOW SLAVE STATUS} on the master at the time
you have discovered the problem
@item
Error logs on the master and on the slave
@end itemize
@item
Use @code{mysqlbinlog} to examine the binary logs. The following should
be helpful
to find the trouble query, for example:
@example
mysqlbinlog -j pos_from_slave_status /path/to/log_from_slave_status | head
@end example
@end itemize

Once you have collected the evidence on the phantom problem, try hard to
isolate it into a separate test case first. Then report the problem to
@email{bugs@@lists.mysql.com} with as much info as possible.




@node MySQL Optimisation, Reference, MySQL Database Administration, Top
@chapter MySQL Optimisation

@menu
* Optimise Overview::           Optimisation Overview
* Query Speed::                 Optimising @code{SELECT}s and Other Queries
* Locking Issues::              Locking Issues
* Optimising Database Structure::  Optimising Database Structure
* Optimising the Server::       Optimising the MySQL Server
* Disk issues::                 Disk Issues
@end menu


Optimisation is a complicated task because it ultimately requires
understanding of the whole system. While it may be possible to do some
local optimisations with small knowledge of your system or application,
the more optimal you want your system to become the more you will have
to know about it.

This chapter will try to explain and give some examples of different
ways to optimise MySQL.  Remember, however, that there are
always some (increasingly harder) additional ways to make the system
even faster.


@node Optimise Overview, Query Speed, MySQL Optimisation, MySQL Optimisation
@section Optimisation Overview

The most important part for getting a system fast is of course the basic
design. You also need to know what kinds of things your system will be
doing, and what your bottlenecks are.

The most common bottlenecks are:
@itemize @bullet
@item Disk seeks.
It takes time for the disk to find a piece of data. With modern disks in
1999, the mean time for this is usually lower than 10ms, so we can in
theory do about 100 seeks a second. This time improves slowly with new
disks and is very hard to optimise for a single table. The way to
optimise this is to spread the data on more than one disk.

@item Disk reading/writing.
When the disk is at the correct position we need to read the data. With
modern disks in 1999, one disk delivers something like 10-20MB/s. This
is easier to optimise than seeks because you can read in parallel from
multiple disks.

@item CPU cycles.
When we have the data in main memory (or if it already were
there) we need to process it to get to our result. Having small
tables compared to the memory is the most common limiting
factor. But then, with small tables speed is usually not the problem.

@item Memory bandwidth.
When the CPU needs more data than can fit in the CPU cache the main
memory bandwidth becomes a bottleneck. This is an uncommon bottleneck
for most systems, but one should be aware of it.
@end itemize


@menu
* Design Limitations::          MySQL Design Limitations/Tradeoffs
* Portability::                 Portability
* Internal use::                What Have We Used MySQL For?
* MySQL Benchmarks::            The MySQL Benchmark Suite
* Custom Benchmarks::           Using Your Own Benchmarks
@end menu


@node Design Limitations, Portability, Optimise Overview, Optimise Overview
@subsection MySQL Design Limitations/Tradeoffs

@cindex design, limitations
@cindex limitations, design

When using the MyISAM table handler, MySQL uses extremely fast table
locking (multiple readers / single writers). The biggest problem with
this table type is a if you have a mix of a steady stream of updates and
slow selects on the same table.  If this is a problem with some tables,
you can use another table type for these. @xref{Table types}.

MySQL can work with both transactional and not transactional tables.  To
be able to work smoothly with not transactional tables (which can't
rollback if something goes wrong), MySQL has the following rules:

@cindex default values
@itemize @bullet
@item
All columns has default values.
@item
If you insert a 'wrong' value in a column like a @code{NULL} in a
@code{NOT NULL} column or a too big numerical value in a numerical
column, MySQL will instead of giving an error instead set the column to
the 'best possible value'.  For numerical values this is 0, the smallest
possible values or the largest possible value. For strings this is
either the empty string or the longest possible string that can be in
the column.
@item
All calculated expressions returns a value that can be used instead of
signaling an error condition. For example 1/0 returns @code{NULL}
@end itemize

The reason for the above rules is that we can't check these conditions
before the query starts to execute.  If we encounter a problem after
updating a few rows, we can't just rollback as the table type may not
support this.  We can't stop because in that case the update would be
'half done' which is probably the worst possible scenario.  In this case
it's better to 'do the best you can' and then continue as if nothing
happened.

The above means that one should not use MySQL to check fields content,
but one should do this in the application.

@node Portability, Internal use, Design Limitations, Optimise Overview
@subsection Portability

@cindex portability
@cindex crash-me program
@cindex programs, crash-me

Because all SQL servers implement different parts of SQL, it takes work to
write portable SQL applications. For very simple selects/inserts it is
very easy, but the more you need the harder it gets. If you want an
application that is fast with many databases it becomes even harder!

To make a complex application portable you need to choose a number of
SQL servers that it should work with.

You can use the MySQL crash-me program/web-page
@uref{http://www.mysql.com/information/crash-me.php} to find functions,
types, and limits you can use with a selection of database
servers. Crash-me now tests far from everything possible, but it
is still comprehensive with about 450 things tested.

For example, you shouldn't have column names longer than 18 characters
if you want to be able to use Informix or DB2.

Both the MySQL benchmarks and crash-me programs are very
database-independent.  By taking a look at how we have handled this, you
can get a feeling for what you have to do to write your application
database-independent.  The benchmarks themselves can be found in the
@file{sql-bench} directory in the MySQL source
distribution. They are written in Perl with DBI database interface
(which solves the access part of the problem).

See @uref{http://www.mysql.com/information/benchmarks.html} for the results
from this benchmark.

As you can see in these results, all databases have some weak points. That
is, they have different design compromises that lead to different
behavior.

If you strive for database independence, you need to get a good feeling
for each SQL server's bottlenecks. MySQL is very fast in
retrieving and updating things, but will have a problem in mixing slow
readers/writers on the same table. Oracle, on the other hand, has a big
problem when you try to access rows that you have recently updated
(until they are flushed to disk). Transaction databases in general are
not very good at generating summary tables from log tables, as in this
case row locking is almost useless.

To get your application @emph{really} database-independent, you need to define
an easy extendable interface through which you manipulate your data. As
C++ is available on most systems, it makes sense to use a C++ classes
interface to the databases.

If you use some specific feature for some database (like the
@code{REPLACE} command in MySQL), you should code a method for
the other SQL servers to implement the same feature (but slower).  With
MySQL you can use the @code{/*!  */} syntax to add
MySQL-specific keywords to a query.  The code inside
@code{/**/} will be treated as a comment (ignored) by most other SQL
servers.

If high performance is more important than exactness, as in some
web applications, it is possibile to create an application layer that
caches all results to give you even higher performance. By letting
old results 'expire' after a while, you can keep the cache reasonably
fresh.  This provides a method to handle high load spikes, in which case
you can dynamically increase the cache and set the expire timeout higher
until things get back to normal.

In this case the table creation information should contain information
of the initial size of the cache and how often the table should normally
be refreshed.


@node Internal use, MySQL Benchmarks, Portability, Optimise Overview
@subsection What Have We Used MySQL For?

@cindex uses, of MySQL
@cindex customers, of MySQL

During MySQL initial development, the features of MySQL
were made to fit our largest customer. They handle data warehousing for a
couple of the biggest retailers in Sweden.

From all stores, we get weekly summaries of all bonus card transactions,
and we are expected to provide useful information for the store owners
to help them find how their advertisement campaigns are affecting their
customers.

The data is quite huge (about 7 million summary transactions per month),
and we have data for 4-10 years that we need to present to the users.
We got weekly requests from the customers that they want to get
'instant' access to new reports from this data.

We solved this by storing all information per month in compressed
'transaction' tables. We have a set of simple macros (script) that
generates summary tables grouped by different criteria (product group,
customer id, store ...) from the transaction tables.  The reports are
web pages that are dynamically generated by a small Perl script that
parses a web page, executes the SQL statements in it, and inserts the
results. We would have used PHP or mod_perl instead but they were
not available at that time.

For graphical data we wrote a simple tool in @code{C} that can produce
GIFs based on the result of a SQL query (with some processing of the
result). This is also dynamically executed from the Perl script that
parses the @code{HTML} files.

In most cases a new report can simply be done by copying an existing
script and modifying the SQL query in it.  In some cases, we will need to
add more fields to an existing summary table or generate a new one, but
this is also quite simple, as we keep all transactions tables on disk.
(Currently we have at least 50G of transactions tables and 200G of other
customer data.)

We also let our customers access the summary tables directly with ODBC
so that the advanced users can themselves experiment with the data.

We haven't had any problems handling this with quite modest Sun Ultra
SPARCstation (2x200 Mhz). We recently upgraded one of our servers to a 2
CPU 400 Mhz UltraSPARC, and we are now planning to start handling
transactions on the product level, which would mean a ten-fold increase
of data. We think we can keep up with this by just adding more disk to
our systems.

We are also experimenting with Intel-Linux to be able to get more CPU
power cheaper. Now that we have the binary portable database format (new
in Version 3.23), we will start to use this for some parts of the application.

Our initial feelings are that Linux will perform much better on
low-to-medium load and Solaris will perform better when you start to get a
high load because of extreme disk IO, but we don't yet have anything
conclusive about this. After some discussion with a Linux Kernel
developer, this might be a side effect of Linux giving so much resources
to the batch job that the interactive performance gets very low. This
makes the machine feel very slow and unresponsive while big batches are
going. Hopefully this will be better handled in future Linux Kernels.


@node MySQL Benchmarks, Custom Benchmarks, Internal use, Optimise Overview
@subsection The MySQL Benchmark Suite

@cindex benchmark suite
@cindex crash-me program

This should contain a technical description of the MySQL
benchmark suite (and @code{crash-me}), but that description is not
written yet. Currently, you can get a good idea of the benchmark by
looking at the code and results in the @file{sql-bench} directory in any
MySQL source distributions.

This benchmark suite is meant to be a benchmark that will tell any user
what things a given SQL implementation performs well or poorly at.

Note that this benchmark is single threaded, so it measures the minimum
time for the operations. We plan to in the future add a lot of
multi-threaded tests to the benchmark suite.

For example, (run on the same NT 4.0 machine):

@multitable @columnfractions .35 .10 .10
@item @strong{Reading 2000000 rows by index} @tab @strong{Seconds} @tab @strong{Seconds}
@item mysql             @tab 367          @tab 249
@item mysql_odbc	@tab 464          @tab
@item db2_odbc		@tab 1206         @tab
@item informix_odbc	@tab 121126       @tab
@item ms-sql_odbc	@tab 1634         @tab
@item oracle_odbc	@tab 20800        @tab
@item solid_odbc	@tab 877          @tab
@item sybase_odbc	@tab 17614        @tab
@end multitable

@multitable @columnfractions .35 .10 .10
@item @strong{Inserting (350768) rows} @tab @strong{Seconds} @tab @strong{Seconds}
@item mysql		@tab 381          @tab 206
@item mysql_odbc	@tab 619          @tab
@item db2_odbc		@tab 3460         @tab
@item informix_odbc	@tab 2692         @tab
@item ms-sql_odbc	@tab 4012         @tab
@item oracle_odbc	@tab 11291        @tab
@item solid_odbc	@tab 1801         @tab
@item sybase_odbc	@tab 4802         @tab
@end multitable

In the above test MySQL was run with a 8M index cache.

We have gathered some more benchmark results at
@uref{http://www.mysql.com/information/benchmarks.html}.

Note that Oracle is not included because they asked to be removed. All
Oracle benchmarks have to be passed by Oracle! We believe that makes
Oracle benchmarks @strong{very} biased because the above benchmarks are
supposed to show what a standard installation can do for a single
client.

To run the benchmark suite, you have to download a MySQL source
distribution, install the perl DBI driver, the perl DBD driver for the
database you want to test and then do:

@example
cd sql-bench
perl run-all-tests --server=#
@end example

where # is one of supported servers. You can get a list of all options
and supported servers by doing @code{run-all-tests --help}.

@cindex crash-me
@code{crash-me} tries to determine what features a database supports and
what its capabilities and limitations are by actually running
queries. For example, it determines:

@itemize @bullet
@item
What column types are supported
@item
How many indexes are supported
@item
What functions are supported
@item
How big a query can be
@item
How big a @code{VARCHAR} column can be
@end itemize

We can find the result from crash-me on a lot of different databases at
@uref{http://www.mysql.com/information/crash-me.php}.


@node Custom Benchmarks,  , MySQL Benchmarks, Optimise Overview
@subsection Using Your Own Benchmarks

@cindex benchmarks
@cindex performance, benchmarks

You should definitely benchmark your application and database to find
out where the bottlenecks are.  By fixing it (or by replacing the
bottleneck with a 'dummy module') you can then easily identify the next
bottleneck (and so on).  Even if the overall performance for your
application is sufficient, you should at least make a plan for each
bottleneck, and decide how to solve it if someday you really need the
extra performance.

For an example of portable benchmark programs, look at the MySQL
benchmark suite. @xref{MySQL Benchmarks, , MySQL Benchmarks}. You
can take any program from this suite and modify it for your needs. By doing
this, you can try different solutions to your problem and test which is really
the fastest solution for you.

It is very common that some problems only occur when the system is very
heavily loaded. We have had many customers who contact us when they
have a (tested) system in production and have encountered load problems. In
every one of these cases so far, it has been problems with basic design
(table scans are @strong{not good} at high load) or OS/Library issues. Most of
this would be a @strong{lot} easier to fix if the systems were not
already in production.

To avoid problems like this, you should put some effort into benchmarking
your whole application under the worst possible load!  You can use
Super Smack for this, and it is available at:
@uref{http://www.mysql.com/Downloads/super-smack/super-smack-1.0.tar.gz}.
As the name suggests, it can bring your system down to its knees if you ask it,
so make sure to use it only on your development systems.


@node Query Speed, Locking Issues, Optimise Overview, MySQL Optimisation
@section Optimising @code{SELECT}s and Other Queries

@cindex queries, speed of
@cindex permission checks, effect on speed
@cindex speed, of queries

First, one thing that affects all queries: The more complex permission
system setup you have, the more overhead you get.

If you do not have any @code{GRANT} statements done, MySQL will
optimise the permission checking somewhat. So if you have a very high
volume it may be worth the time to avoid grants. Otherwise, more
permission check results in a larger overhead.

If your problem is with some explicit MySQL function, you can
always time this in the MySQL client:

@example
mysql> SELECT BENCHMARK(1000000,1+1);
+------------------------+
| BENCHMARK(1000000,1+1) |
+------------------------+
|                      0 |
+------------------------+
1 row in set (0.32 sec)
@end example

The above shows that MySQL can execute 1,000,000 @code{+}
expressions in 0.32 seconds on a @code{PentiumII 400MHz}.

All MySQL functions should be very optimised, but there may be
some exceptions, and the @code{BENCHMARK(loop_count,expression)} is a
great tool to find out if this is a problem with your query.

@menu
* EXPLAIN::                     @code{EXPLAIN} Syntax (Get Information About a @code{SELECT})
* Estimating performance::      Estimating query performance
* SELECT speed::                Speed of @code{SELECT} queries
* Where optimisations::         How MySQL optimises @code{WHERE} clauses
* DISTINCT optimisation::       How MySQL Optimises @code{DISTINCT}
* LEFT JOIN optimisation::      How MySQL optimises @code{LEFT JOIN}
* ORDER BY optimisation::       How MySQL Optimises @code{ORDER BY}
* LIMIT optimisation::          How MySQL optimises @code{LIMIT}
* Insert speed::                Speed of @code{INSERT} queries
* Update speed::                Speed of @code{UPDATE} queries
* Delete speed::                Speed of @code{DELETE} queries
* Tips::                        Other Optimisation Tips
@end menu


@node EXPLAIN, Estimating performance, Query Speed, Query Speed
@subsection @code{EXPLAIN} Syntax (Get Information About a @code{SELECT})

@findex EXPLAIN
@findex SELECT, optimising

@example
    EXPLAIN tbl_name
or  EXPLAIN SELECT select_options
@end example

@code{EXPLAIN tbl_name} is a synonym for @code{DESCRIBE tbl_name} or
@code{SHOW COLUMNS FROM tbl_name}.

When you precede a @code{SELECT} statement with the keyword @code{EXPLAIN},
MySQL explains how it would process the @code{SELECT}, providing
information about how tables are joined and in which order.

With the help of @code{EXPLAIN}, you can see when you must add indexes
to tables to get a faster @code{SELECT} that uses indexes to find the
records. You can also see if the optimiser joins the tables in an optimal
order. To force the optimiser to use a specific join order for a
@code{SELECT} statement, add a @code{STRAIGHT_JOIN} clause.

For non-simple joins, @code{EXPLAIN} returns a row of information for each
table used in the @code{SELECT} statement. The tables are listed in the order
they would be read.  MySQL resolves all joins using a single-sweep
multi-join method. This means that MySQL reads a row from the first
table, then finds a matching row in the second table, then in the third table
and so on. When all tables are processed, it outputs the selected columns and
backtracks through the table list until a table is found for which there are
more matching rows. The next row is read from this table and the process
continues with the next table.

Output from @code{EXPLAIN} includes the following columns:

@table @code
@item table
The table to which the row of output refers.

@item type
The join type.  Information about the various types is given below.

@item possible_keys
The @code{possible_keys} column indicates which indexes MySQL
could use to find the rows in this table. Note that this column is
totally independent of the order of the tables. That means that some of
the keys in possible_keys may not be usable in practice with the
generated table order.

If this column is empty, there are no relevant indexes. In this case,
you may be able to improve the performance of your query by examining
the @code{WHERE} clause to see if it refers to some column or columns
that would be suitable for indexing.  If so, create an appropriate index
and check the query with @code{EXPLAIN} again. @xref{ALTER TABLE}.

To see what indexes a table has, use @code{SHOW INDEX FROM tbl_name}.

@item key
The @code{key} column indicates the key (index) that MySQL actually
decided to use. The key is @code{NULL} if no index was chosen.  If
MySQL chooses the wrong index, you can probably force
MySQL to use another index by using @code{myisamchk --analyze},
@xref{myisamchk syntax}, or by using @code{USE INDEX/IGNORE INDEX}.
@xref{SELECT}.

@item key_len
The @code{key_len} column indicates the length of the key that
MySQL decided to use.  The length is @code{NULL} if the
@code{key} is @code{NULL}. Note that this tells us how many parts of a
multi-part key MySQL will actually use.

@item ref
The @code{ref} column shows which columns or constants are used with the
@code{key} to select rows from the table.

@item rows
The @code{rows} column indicates the number of rows MySQL
believes it must examine to execute the query.

@item Extra
This column contains additional information of how MySQL will
resolve the query. Here is an explanation of the different text
strings that can be found in this column:

@table @code
@item Distinct
MySQL will not continue searching for more rows for the current row
combination after it has found the first matching row.

@item Not exists
MySQL was able to do a @code{LEFT JOIN} optimisation on the
query and will not examine more rows in this table for the previous row
combination after it finds one row that matches the @code{LEFT JOIN} criteria.

Here is an example for this:

@example
SELECT * FROM t1 LEFT JOIN t2 ON t1.id=t2.id WHERE t2.id IS NULL;
@end example

Assume that @code{t2.id} is defined with @code{NOT NULL}.  In this case
MySQL will scan @code{t1} and look up the rows in @code{t2}
through @code{t1.id}. If MySQL finds a matching row in
@code{t2}, it knows that @code{t2.id} can never be @code{NULL}, and will
not scan through the rest of the rows in @code{t2} that has the same
@code{id}.  In other words, for each row in @code{t1}, MySQL
only needs to do a single lookup in @code{t2}, independent of how many
matching rows there are in @code{t2}.

@item @code{range checked for each record (index map: #)}
MySQL didn't find a real good index to use. It will, instead, for
each row combination in the preceding tables, do a check on which index to
use (if any), and use this index to retrieve the rows from the table.  This
isn't very fast but is faster than having to do a join without
an index.

@item Using filesort
MySQL will need to do an extra pass to find out how to retrieve
the rows in sorted order.  The sort is done by going through all rows
according to the @code{join type} and storing the sort key + pointer to
the row for all rows that match the @code{WHERE}. Then the keys are
sorted. Finally the rows are retrieved in sorted order.

@item Using index
The column information is retrieved from the table using only
information in the index tree without having to do an additional seek to
read the actual row.  This can be done when all the used columns for
the table are part of the same index.

@item Using temporary
To resolve the query MySQL will need to create a
temporary table to hold the result.  This typically happens if you do an
@code{ORDER BY} on a different column set than you did a @code{GROUP
BY} on.

@item Where used
A @code{WHERE} clause will be used to restrict which rows will be
matched against the next table or sent to the client.  If you don't have
this information and the table is of type @code{ALL} or @code{index},
you may have something wrong in your query (if you don't intend to
fetch/examine all rows from the table).
@end table

If you want to get your queries as fast as possible, you should look out for
@code{Using filesort} and @code{Using temporary}.
@end table

The different join types are listed here, ordered from best to worst type:

@cindex system table
@cindex tables, system
@table @code
@item system
The table has only one row (= system table). This is a special case of
the @code{const} join type.

@cindex constant table
@cindex tables, constant
@item const
The table has at most one matching row, which will be read at the start
of the query. Because there is only one row, values from the column in
this row can be regarded as constants by the rest of the
optimiser. @code{const} tables are very fast as they are read only once!

@item eq_ref
One row will be read from this table for each combination of rows from
the previous tables.  This is the best possible join type, other than the
@code{const} types.  It is used when all parts of an index are used by
the join and the index is @code{UNIQUE} or a @code{PRIMARY KEY}.

@item ref
All rows with matching index values will be read from this table for each
combination of rows from the previous tables.  @code{ref} is used if the join
uses only a leftmost prefix of the key, or if the key is not @code{UNIQUE}
or a @code{PRIMARY KEY} (in other words, if the join cannot select a single
row based on the key value).  If the key that is used matches only a few rows,
this join type is good.

@item range
Only rows that are in a given range will be retrieved, using an index to
select the rows.  The @code{key} column indicates which index is used.
The @code{key_len} contains the longest key part that was used.
The @code{ref} column will be NULL for this type.

@item index
This is the same as @code{ALL}, except that only the index tree is
scanned.  This is usually faster than @code{ALL}, as the index file is usually
smaller than the datafile.

@item ALL
A full table scan will be done for each combination of rows from the
previous tables.  This is normally not good if the table is the first
table not marked @code{const}, and usually @strong{very} bad in all other
cases. You normally can avoid @code{ALL} by adding more indexes, so that
the row can be retrieved based on constant values or column values from
earlier tables.
@end table

You can get a good indication of how good a join is by multiplying all values
in the @code{rows} column of the @code{EXPLAIN} output. This should tell you
roughly how many rows MySQL must examine to execute the query. This
number is also used when you restrict queries with the @code{max_join_size}
variable.
@xref{Server parameters}.

The following example shows how a @code{JOIN} can be optimised progressively
using the information provided by @code{EXPLAIN}.

Suppose you have the @code{SELECT} statement shown here, that you examine
using @code{EXPLAIN}:

@example
EXPLAIN SELECT tt.TicketNumber, tt.TimeIn,
            tt.ProjectReference, tt.EstimatedShipDate,
            tt.ActualShipDate, tt.ClientID,
            tt.ServiceCodes, tt.RepetitiveID,
            tt.CurrentProcess, tt.CurrentDPPerson,
            tt.RecordVolume, tt.DPPrinted, et.COUNTRY,
            et_1.COUNTRY, do.CUSTNAME
        FROM tt, et, et AS et_1, do
        WHERE tt.SubmitTime IS NULL
            AND tt.ActualPC = et.EMPLOYID
            AND tt.AssignedPC = et_1.EMPLOYID
            AND tt.ClientID = do.CUSTNMBR;
@end example

For this example, assume that:

@itemize @bullet
@item
The columns being compared have been declared as follows:

@multitable @columnfractions .10 .15 .15
@item @strong{Table} @tab @strong{Column} @tab @strong{Column type}
@item @code{tt}      @tab @code{ActualPC}      @tab @code{CHAR(10)}
@item @code{tt}      @tab @code{AssignedPC}    @tab @code{CHAR(10)}
@item @code{tt}      @tab @code{ClientID}      @tab @code{CHAR(10)}
@item @code{et}      @tab @code{EMPLOYID}      @tab @code{CHAR(15)}
@item @code{do}      @tab @code{CUSTNMBR}      @tab @code{CHAR(15)}
@end multitable

@item
The tables have the indexes shown here:

@multitable @columnfractions .10 .30
@item @strong{Table} @tab @strong{Index}
@item @code{tt}      @tab @code{ActualPC}
@item @code{tt}      @tab @code{AssignedPC}
@item @code{tt}      @tab @code{ClientID}
@item @code{et}      @tab @code{EMPLOYID} (primary key)
@item @code{do}      @tab @code{CUSTNMBR} (primary key)
@end multitable

@item
The @code{tt.ActualPC} values aren't evenly distributed.
@end itemize

Initially, before any optimisations have been performed, the @code{EXPLAIN}
statement produces the following information:

@example
table type possible_keys                key  key_len ref  rows  Extra
et    ALL  PRIMARY                      NULL NULL    NULL 74
do    ALL  PRIMARY                      NULL NULL    NULL 2135
et_1  ALL  PRIMARY                      NULL NULL    NULL 74
tt    ALL  AssignedPC,ClientID,ActualPC NULL NULL    NULL 3872
      range checked for each record (key map: 35)
@end example

Because @code{type} is @code{ALL} for each table, this output indicates that
MySQL is doing a full join for all tables!  This will take quite a
long time, as the product of the number of rows in each table must be
examined!  For the case at hand, this is @code{74 * 2135 * 74 * 3872 =
45,268,558,720} rows.  If the tables were bigger, you can only imagine how
long it would take.

One problem here is that MySQL can't (yet) use indexes on columns
efficiently if they are declared differently.  In this context,
@code{VARCHAR} and @code{CHAR} are the same unless they are declared as
different lengths. Because @code{tt.ActualPC} is declared as @code{CHAR(10)}
and @code{et.EMPLOYID} is declared as @code{CHAR(15)}, there is a length
mismatch.

To fix this disparity between column lengths, use @code{ALTER TABLE} to
lengthen @code{ActualPC} from 10 characters to 15 characters:

@example
mysql> ALTER TABLE tt MODIFY ActualPC VARCHAR(15);
@end example

Now @code{tt.ActualPC} and @code{et.EMPLOYID} are both @code{VARCHAR(15)}.
Executing the @code{EXPLAIN} statement again produces this result:

@example
table type   possible_keys   key     key_len ref         rows    Extra
tt    ALL    AssignedPC,ClientID,ActualPC NULL NULL NULL 3872    where used
do    ALL    PRIMARY         NULL    NULL    NULL        2135
      range checked for each record (key map: 1)
et_1  ALL    PRIMARY         NULL    NULL    NULL        74
      range checked for each record (key map: 1)
et    eq_ref PRIMARY         PRIMARY 15      tt.ActualPC 1
@end example

This is not perfect, but is much better (the product of the @code{rows}
values is now less by a factor of 74). This version is executed in a couple
of seconds.

A second alteration can be made to eliminate the column length mismatches
for the @code{tt.AssignedPC = et_1.EMPLOYID} and @code{tt.ClientID =
do.CUSTNMBR} comparisons:

@example
mysql> ALTER TABLE tt MODIFY AssignedPC VARCHAR(15),
    ->                MODIFY ClientID   VARCHAR(15);
@end example

Now @code{EXPLAIN} produces the output shown here:

@example
table type   possible_keys   key      key_len ref           rows Extra
et    ALL    PRIMARY         NULL     NULL    NULL          74
tt    ref    AssignedPC,     ActualPC 15      et.EMPLOYID   52   where used
             ClientID,
             ActualPC
et_1  eq_ref PRIMARY         PRIMARY  15      tt.AssignedPC 1
do    eq_ref PRIMARY         PRIMARY  15      tt.ClientID   1
@end example

This is almost as good as it can get.

The remaining problem is that, by default, MySQL assumes that values
in the @code{tt.ActualPC} column are evenly distributed, and that isn't the
case for the @code{tt} table.  Fortunately, it is easy to tell MySQL
about this:

@example
shell> myisamchk --analyze PATH_TO_MYSQL_DATABASE/tt
shell> mysqladmin refresh
@end example

Now the join is perfect, and @code{EXPLAIN} produces this result:

@example
table type   possible_keys key     key_len ref           rows Extra
tt    ALL    AssignedPC    NULL    NULL    NULL          3872 where used
             ClientID,
             ActualPC
et    eq_ref PRIMARY       PRIMARY 15      tt.ActualPC   1
et_1  eq_ref PRIMARY       PRIMARY 15      tt.AssignedPC 1
do    eq_ref PRIMARY       PRIMARY 15      tt.ClientID   1
@end example

Note that the @code{rows} column in the output from @code{EXPLAIN} is an
educated guess from the MySQL join optimiser. To optimise a
query, you should check if the numbers are even close to the truth.  If not,
you may get better performance by using @code{STRAIGHT_JOIN} in your
@code{SELECT} statement and trying to list the tables in a different order in
the @code{FROM} clause.


@node Estimating performance, SELECT speed, EXPLAIN, Query Speed
@subsection Estimating Query Performance

@cindex estimating, query performance
@cindex queries, estimating performance
@cindex performance, estimating

In most cases you can estimate the performance by counting disk seeks.
For small tables, you can usually find the row in 1 disk seek (as the
index is probably cached).  For bigger tables, you can estimate that
(using B++ tree indexes) you will need: @code{log(row_count) /
log(index_block_length / 3 * 2 / (index_length + data_pointer_length)) +
1} seeks to find a row.

In MySQL an index block is usually 1024 bytes and the data
pointer is usually 4 bytes. A 500,000 row table with an
index length of 3 (medium integer) gives you:
@code{log(500,000)/log(1024/3*2/(3+4)) + 1} = 4 seeks.

As the above index would require about 500,000 * 7 * 3/2 = 5.2M,
(assuming that the index buffers are filled to 2/3, which is typical)
you will probably have much of the index in memory and you will probably
only need 1-2 calls to read data from the OS to find the row.

For writes, however, you will need 4 seek requests (as above) to find
where to place the new index and normally 2 seeks to update the index
and write the row.

Note that the above doesn't mean that your application will slowly
degenerate by N log N!  As long as everything is cached by the OS or SQL
server things will only go marginally slower while the table gets
bigger. After the data gets too big to be cached, things will start to
go much slower until your applications is only bound by disk-seeks
(which increase by N log N). To avoid this, increase the index cache as
the data grows. @xref{Server parameters}.


@node SELECT speed, Where optimisations, Estimating performance, Query Speed
@subsection Speed of @code{SELECT} Queries

@findex SELECT speed

@cindex speed, of queries

In general, when you want to make a slow @code{SELECT ... WHERE} faster, the
first thing to check is whether you can add an index. @xref{MySQL
indexes, , MySQL indexes}. All references between different tables
should usually be done with indexes. You can use the @code{EXPLAIN} command
to determine which indexes are used for a @code{SELECT}.
@xref{EXPLAIN, , @code{EXPLAIN}}.

Some general tips:

@itemize @bullet
@item
To help MySQL optimise queries better, run @code{myisamchk
--analyze} on a table after it has been loaded with relevant data. This
updates a value for each index part that indicates the average number of
rows that have the same value.  (For unique indexes, this is always 1,
of course.)  MySQL will use this to decide which index to
choose when you connect two tables with 'a non-constant expression'.
You can check the result from the @code{analyze} run by doing @code{SHOW
INDEX FROM table_name} and examining the @code{Cardinality} column.

@item
To sort an index and data according to an index, use @code{myisamchk
--sort-index --sort-records=1} (if you want to sort on index 1). If you
have a unique index from which you want to read all records in order
according to that index, this is a good way to make that faster.  Note,
however, that this sorting isn't written optimally and will take a long
time for a large table!
@end itemize


@node Where optimisations, DISTINCT optimisation, SELECT speed, Query Speed
@subsection How MySQL Optimises @code{WHERE} Clauses

@findex WHERE

@cindex optimisations

The @code{WHERE} optimisations are put in the @code{SELECT} part here because
they are mostly used with @code{SELECT}, but the same optimisations apply for
@code{WHERE} in @code{DELETE} and @code{UPDATE} statements.

Also note that this section is incomplete. MySQL does many
optimisations, and we have not had time to document them all.

Some of the optimisations performed by MySQL are listed here:

@itemize @bullet
@item
Removal of unnecessary parentheses:
@example
   ((a AND b) AND c OR (((a AND b) AND (c AND d))))
-> (a AND b AND c) OR (a AND b AND c AND d)
@end example
@item
Constant folding:
@example
   (a<b AND b=c) AND a=5
-> b>5 AND b=c AND a=5
@end example
@item
Constant condition removal (needed because of constant folding):
@example
   (B>=5 AND B=5) OR (B=6 AND 5=5) OR (B=7 AND 5=6)
-> B=5 OR B=6
@end example
@item
Constant expressions used by indexes are evaluated only once.
@item
@code{COUNT(*)} on a single table without a @code{WHERE} is retrieved
directly from the table information for MyISAM and HEAP tables.  This is
also done for any @code{NOT NULL} expression when used with only one table.
@item
Early detection of invalid constant expressions. MySQL quickly
detects that some @code{SELECT} statements are impossible and returns no rows.
@item
@code{HAVING} is merged with @code{WHERE} if you don't use @code{GROUP BY}
or group functions (@code{COUNT()}, @code{MIN()}...).
@item
For each sub-join, a simpler @code{WHERE} is constructed to get a fast
@code{WHERE} evaluation for each sub-join and also to skip records as
soon as possible.
@cindex constant table
@cindex tables, constant
@item
All constant tables are read first, before any other tables in the query.
A constant table is:
@itemize @minus
@item
An empty table or a table with 1 row.
@item
A table that is used with a @code{WHERE} clause on a @code{UNIQUE}
index, or a @code{PRIMARY KEY}, where all index parts are used with constant
expressions and the index parts are defined as @code{NOT NULL}.
@end itemize
All the following tables are used as constant tables:
@example
mysql> SELECT * FROM t WHERE primary_key=1;
mysql> SELECT * FROM t1,t2
    ->          WHERE t1.primary_key=1 AND t2.primary_key=t1.id;
@end example

@item
The best join combination to join the tables is found by trying all
possibilities. If all columns in @code{ORDER BY} and in @code{GROUP
BY} come from the same table, then this table is preferred first when
joining.
@item
If there is an @code{ORDER BY} clause and a different @code{GROUP BY}
clause, or if the @code{ORDER BY} or @code{GROUP BY} contains columns
from tables other than the first table in the join queue, a temporary
table is created.
@item
If you use @code{SQL_SMALL_RESULT}, MySQL will use an in-memory
temporary table.
@item
Each table index is queried, and the best index that spans fewer than 30% of
the rows is used. If no such index can be found, a quick table scan is used.
@item
In some cases, MySQL can read rows from the index without even
consulting the datafile.  If all columns used from the index are numeric,
then only the index tree is used to resolve the query.
@item
Before each record is output, those that do not match the @code{HAVING} clause
are skipped.
@end itemize

Some examples of queries that are very fast:

@example
mysql> SELECT COUNT(*) FROM tbl_name;
mysql> SELECT MIN(key_part1),MAX(key_part1) FROM tbl_name;
mysql> SELECT MAX(key_part2) FROM tbl_name
    ->        WHERE key_part_1=constant;
mysql> SELECT ... FROM tbl_name
    ->        ORDER BY key_part1,key_part2,... LIMIT 10;
mysql> SELECT ... FROM tbl_name
    ->        ORDER BY key_part1 DESC,key_part2 DESC,... LIMIT 10;
@end example

The following queries are resolved using only the index tree (assuming
the indexed columns are numeric):

@example
mysql> SELECT key_part1,key_part2 FROM tbl_name WHERE key_part1=val;
mysql> SELECT COUNT(*) FROM tbl_name
    ->        WHERE key_part1=val1 AND key_part2=val2;
mysql> SELECT key_part2 FROM tbl_name GROUP BY key_part1;
@end example

The following queries use indexing to retrieve the rows in sorted
order without a separate sorting pass:

@example
mysql> SELECT ... FROM tbl_name
    ->            ORDER BY key_part1,key_part2,... ;
mysql> SELECT ... FROM tbl_name
    ->            ORDER BY key_part1 DESC,key_part2 DESC,... ;
@end example


@node DISTINCT optimisation, LEFT JOIN optimisation, Where optimisations, Query Speed
@subsection How MySQL Optimises @code{DISTINCT}

@findex DISTINCT

@cindex optimising, DISTINCT

@code{DISTINCT} is converted to a @code{GROUP BY} on all columns,
@code{DISTINCT} combined with @code{ORDER BY} will in many cases also
need a temporary table.

When combining @code{LIMIT #} with @code{DISTINCT}, MySQL will stop
as soon as it finds @code{#} unique rows.

If you don't use columns from all used tables, MySQL will stop
the scanning of the not used tables as soon as it has found the first match.

@example
SELECT DISTINCT t1.a FROM t1,t2 where t1.a=t2.a;
@end example

In the case, assuming t1 is used before t2 (check with @code{EXPLAIN}), then
MySQL will stop reading from t2 (for that particular row in t1)
when the first row in t2 is found.


@node LEFT JOIN optimisation, ORDER BY optimisation, DISTINCT optimisation, Query Speed
@subsection How MySQL Optimises @code{LEFT JOIN} and @code{RIGHT JOIN}

@findex LEFT JOIN

@cindex optimising, LEFT JOIN

@code{A LEFT JOIN B} in MySQL is implemented as follows:

@itemize @bullet
@item
The table @code{B} is set to be dependent on table @code{A} and all tables
that @code{A} is dependent on.

@item
The table @code{A} is set to be dependent on all tables (except @code{B})
that are used in the @code{LEFT JOIN} condition.

@item
All @code{LEFT JOIN} conditions are moved to the @code{WHERE} clause.

@item
All standard join optimisations are done, with the exception that a table is
always read after all tables it is dependent on.  If there is a circular
dependence then MySQL will issue an error.

@item
All standard @code{WHERE} optimisations are done.

@item
If there is a row in @code{A} that matches the @code{WHERE} clause, but there
wasn't any row in @code{B} that matched the @code{LEFT JOIN} condition,
then an extra @code{B} row is generated with all columns set to @code{NULL}.

@item
If you use @code{LEFT JOIN} to find rows that don't exist in some
table and you have the following test: @code{column_name IS NULL} in the
@code{WHERE} part, where column_name is a column that is declared as
@code{NOT NULL}, then MySQL will stop searching after more rows
(for a particular key combination) after it has found one row that
matches the @code{LEFT JOIN} condition.
@end itemize

@code{RIGHT JOIN} is implemented analogously as @code{LEFT JOIN}.

The table read order forced by @code{LEFT JOIN} and @code{STRAIGHT JOIN}
will help the join optimiser (which calculates in which order tables
should be joined) to do its work much more quickly, as there are fewer
table permutations to check.

Note that the above means that if you do a query of type:

@example
SELECT * FROM a,b LEFT JOIN c ON (c.key=a.key) LEFT JOIN d (d.key=a.key)
         WHERE b.key=d.key
@end example

MySQL will do a full scan on @code{b} as the @code{LEFT JOIN} will force
it to be read before @code{d}.

The fix in this case is to change the query to:

@example
SELECT * FROM b,a LEFT JOIN c ON (c.key=a.key) LEFT JOIN d (d.key=a.key)
         WHERE b.key=d.key
@end example


@node ORDER BY optimisation, LIMIT optimisation, LEFT JOIN optimisation, Query Speed
@subsection How MySQL Optimises @code{ORDER BY}

In some cases MySQL can uses index to satisfy an @code{ORDER BY} or
@code{GROUP BY} request without doing any extra sorting.

The index can also be used even if the @code{ORDER BY} doesn't match the
index exactly, as long as all the unused index parts and all the extra
are @code{ORDER BY} columns are constants in the @code{WHERE}
clause. The following queries will use the index to resolve the
@code{ORDER BY} / @code{GROUP BY} part:

@example
SELECT * FROM t1 ORDER BY key_part1,key_part2,...
SELECT * FROM t1 WHERE key_part1=constant ORDER BY key_part2
SELECT * FROM t1 WHERE key_part1=constant GROUP BY key_part2
SELECT * FROM t1 ORDER BY key_part1 DESC,key_part2 DESC
SELECT * FROM t1 WHERE key_part1=1 ORDER BY key_part1 DESC,key_part2 DESC
@end example

Some cases where MySQL can @strong{not} use indexes to resolve the @code{ORDER
BY}: (Note that MySQL will still use indexes to find the rows that
matches the @code{WHERE} clause):

@itemize @bullet
@item
You are doing an @code{ORDER BY} on different keys:

@code{SELECT * FROM t1 ORDER BY key1,key2}
@item
You are doing an @code{ORDER BY} using non-consecutive key parts.

@code{SELECT * FROM t1 WHERE key2=constant ORDER BY key_part2}

@item
You are mixing @code{ASC} and @code{DESC}.

@code{SELECT * FROM t1 ORDER BY key_part1 DESC,key_part2 ASC}

@item
The key used to fetch the rows are not the same one that is used to
do the @code{ORDER BY}:

@code{SELECT * FROM t1 WHERE key2=constant ORDER BY key1}

@item
You are joining many tables and the columns you are doing an @code{ORDER
BY} on are not all from the first not-const table that is used to
retrieve rows (This is the first table in the @code{EXPLAIN} output which
doesn't use a @code{const} row fetch method).

@item
You have different @code{ORDER BY} and @code{GROUP BY} expressions.

@item
The used table index is an index type that doesn't store rows in order.
(Like the @code{HASH} index in @code{HEAP} tables).

@item
The index colum may contain @code{NULL} values and one is using
@code{ORDER BY ... DESC}.  This is because in SQL @code{NULL} values is
always sorted before normal values, independent of you are using
@code{DESC} or not.
@end itemize


In the cases where MySQL have to sort the result, it uses the following
algorithm:

@itemize @bullet
@item
Read all rows according to key or by table scanning.
Rows that doesn't match the WHERE clause are skipped.
@item
Store the sort-key in a buffer (of size @code{sort_buffer}).
@item
When the buffer gets full, run a qsort on it and store the result
in a temporary file.  Save a pointer to the sorted block.
(In the case where all rows fits into the sort buffer, no temporary
file is created)
@item
Repeat the above until all rows have been read.
@item
Do a multi-merge of up to @code{MERGEBUFF} (7) regions to one block in
another temporary file.  Repeat until all blocks from the first file
are in the second file.
@item
Repeat the following until there is less than @code{MERGEBUFF2} (15)
blocks left.
@item
On the last multi-merge, only the pointer to the row (last part of
the sort-key) is written to a result file.
@item
Now the code in @file{sql/records.cc} will be used to read through them
in sorted order by using the row pointers in the result file.  To
optimize this, we read in a big block of row pointers, sort these and
then we read the rows in the sorted order into a row buffer
(@code{record_rnd_buffer}) .
@end itemize

You can with @code{EXPLAIN SELECT ... ORDER BY} check if MySQL can use
indexes to resolve the query.  If you get @code{Using filesort} in the
@code{extra} column, then MySQL can't use indexes to resolve the
@code{ORDER BY}. @xref{EXPLAIN}.

If you want to have a higher @code{ORDER BY} speed, you should first
see if you can get MySQL to use indexes instead of having to do an extra
sorting phase. If this is not possible, then you can do:

@itemize @bullet
@item
Increase the size of the @code{sort_buffer} variable.
@item
Increase the size of the @code{record_rnd_buffer} variable.
@item
Change @code{tmpdir} to point to a dedicated disk with lots of empty space.
@end itemize

@node LIMIT optimisation, Insert speed, ORDER BY optimisation, Query Speed
@subsection How MySQL Optimises @code{LIMIT}

@findex LIMIT

@cindex optimising, LIMIT

In some cases MySQL will handle the query differently when you are
using @code{LIMIT #} and not using @code{HAVING}:

@itemize @bullet
@item
If you are selecting only a few rows with @code{LIMIT}, MySQL
will use indexes in some cases when it normally would prefer to do a
full table scan.
@item
If you use @code{LIMIT #} with @code{ORDER BY}, MySQL will end the
sorting as soon as it has found the first @code{#} lines instead of sorting
the whole table.
@item
When combining @code{LIMIT #} with @code{DISTINCT}, MySQL will stop
as soon as it finds @code{#} unique rows.
@item
In some cases a @code{GROUP BY} can be resolved by reading the key in order
(or do a sort on the key) and then calculate summaries until the
key value changes.  In this case @code{LIMIT #} will not calculate any
unnecessary @code{GROUP BY}'s.
@item
As soon as MySQL has sent the first @code{#} rows to the client, it
will abort the query (If you are not using @code{SQL_CALC_FOUND_ROWS}).
@item
@code{LIMIT 0} will always quickly return an empty set.  This is useful
to check the query and to get the column types of the result columns.
@item
When the server uses temporary tables to resolve the query, the
@code{LIMIT #} is used to calculate how much space is required.
@end itemize


@node Insert speed, Update speed, LIMIT optimisation, Query Speed
@subsection Speed of @code{INSERT} Queries

@findex INSERT

@cindex speed, inserting
@cindex inserting, speed of

The time to insert a record consists approximately of:

@itemize @bullet
@item
Connect:                 (3)
@item
Sending query to server: (2)
@item
Parsing query:           (2)
@item
Inserting record:        (1 x size of record)
@item
Inserting indexes:       (1 x number of indexes)
@item
Close:                   (1)
@end itemize

where the numbers are somewhat proportional to the overall time. This
does not take into consideration the initial overhead to open tables
(which is done once for each concurrently running query).

The size of the table slows down the insertion of indexes by N log N
(B-trees).

Some ways to speed up inserts:

@itemize @bullet
@item
If you are inserting many rows from the same client at the same time, use
multiple value lists @code{INSERT} statements. This is much faster (many times
in some cases) than using separate @code{INSERT} statements.  If you are adding
data to non-empty table, you may tune up @code{myisam_bulk_insert_tree_size}
variable to make it even faster. @xref{SHOW VARIABLES}.
@item
If you are inserting a lot of rows from different clients, you can get
higher speed by using the @code{INSERT DELAYED} statement. @xref{INSERT,
, @code{INSERT}}.
@item
Note that with @code{MyISAM} tables you can insert rows at the same time
@code{SELECT}s are running if there are no deleted rows in the tables.
@item
When loading a table from a text file, use @code{LOAD DATA INFILE}. This
is usually 20 times faster than using a lot of @code{INSERT} statements.
@xref{LOAD DATA, , @code{LOAD DATA}}.
@item
It is possible with some extra work to make @code{LOAD DATA INFILE} run even
faster when the table has many indexes. Use the following procedure:

@enumerate
@item
Optionally create the table with @code{CREATE TABLE}. For example, using
@code{mysql} or Perl-DBI.

@item
Execute a @code{FLUSH TABLES} statement or the shell command @code{mysqladmin
flush-tables}.

@item
Use @code{myisamchk --keys-used=0 -rq /path/to/db/tbl_name}. This will
remove all usage of all indexes from the table.

@item
Insert data into the table with @code{LOAD DATA INFILE}. This will not
update any indexes and will therefore be very fast.

@item
If you are going to only read the table in the future, run @code{myisampack}
on it to make it smaller. @xref{Compressed format}.

@item
Re-create the indexes with @code{myisamchk -r -q
/path/to/db/tbl_name}. This will create the index tree in memory before
writing it to disk, which is much faster because it avoids lots of disk
seeks. The resulting index tree is also perfectly balanced.

@item
Execute a @code{FLUSH TABLES} statement or the shell command @code{mysqladmin
flush-tables}.
@end enumerate

Note that @code{LOAD DATA INFILE} also does the above optimization if
you insert into an empty table; the main difference with the above
procedure is that you can let myisamchk allocate much more temporary
memory for the index creation that you may want MySQL to allocate for
every index recreation.

Since @strong{MySQL 4.0} you can also use
@code{ALTER TABLE tbl_name DISABLE KEYS} instead of
@code{myisamchk --keys-used=0 -rq /path/to/db/tbl_name} and
@code{ALTER TABLE tbl_name ENABLE KEYS} instead of
@code{myisamchk -r -q /path/to/db/tbl_name}. This way you can also skip
@code{FLUSH TABLES} steps.
@item
You can speed up insertions that is done over multiple statements by
locking your tables:

@example
mysql> LOCK TABLES a WRITE;
mysql> INSERT INTO a VALUES (1,23),(2,34),(4,33);
mysql> INSERT INTO a VALUES (8,26),(6,29);
mysql> UNLOCK TABLES;
@end example

The main speed difference is that the index buffer is flushed to disk only
once, after all @code{INSERT} statements have completed. Normally there would
be as many index buffer flushes as there are different @code{INSERT}
statements. Locking is not needed if you can insert all rows with a single
statement.

For transactional tables, you should use @code{BEGIN/COMMIT} instead of
@code{LOCK TABLES} to get a speedup.

Locking will also lower the total time of multi-connection tests, but the
maximum wait time for some threads will go up (because they wait for
locks).  For example:

@example
thread 1 does 1000 inserts
thread 2, 3, and 4 does 1 insert
thread 5 does 1000 inserts
@end example

If you don't use locking, 2, 3, and 4 will finish before 1 and 5. If you
use locking, 2, 3, and 4 probably will not finish before 1 or 5, but the
total time should be about 40% faster.

As @code{INSERT}, @code{UPDATE}, and @code{DELETE} operations are very
fast in MySQL, you will obtain better overall performance by
adding locks around everything that does more than about 5 inserts or
updates in a row.  If you do very many inserts in a row, you could do a
@code{LOCK TABLES} followed by an @code{UNLOCK TABLES} once in a while
(about each 1000 rows) to allow other threads access to the table. This
would still result in a nice performance gain.

Of course, @code{LOAD DATA INFILE} is much faster for loading data.
@end itemize

To get some more speed for both @code{LOAD DATA INFILE} and
@code{INSERT}, enlarge the key buffer. @xref{Server parameters}.


@node Update speed, Delete speed, Insert speed, Query Speed
@subsection Speed of @code{UPDATE} Queries

Update queries are optimised as a @code{SELECT} query with the additional
overhead of a write. The speed of the write is dependent on the size of
the data that is being updated and the number of indexes that are
updated.  Indexes that are not changed will not be updated.

Also, another way to get fast updates is to delay updates and then do
many updates in a row later. Doing many updates in a row is much quicker
than doing one at a time if you lock the table.

Note that, with dynamic record format, updating a record to
a longer total length may split the record.  So if you do this often,
it is very important to @code{OPTIMIZE TABLE} sometimes.
@xref{OPTIMIZE TABLE, , @code{OPTIMIZE TABLE}}.


@node Delete speed, Tips, Update speed, Query Speed
@subsection Speed of @code{DELETE} Queries

If you want to delete all rows in the table, you should use
@code{TRUNCATE TABLE table_name}. @xref{TRUNCATE}.

The time to delete a record is exactly proportional to the number of
indexes. To delete records more quickly, you can increase the size of
the index cache. @xref{Server parameters}.


@node Tips,  , Delete speed, Query Speed
@subsection Other Optimisation Tips

@cindex optimisation, tips
@cindex tips, optimisation

Unsorted tips for faster systems:

@itemize @bullet
@item
Use persistent connections to the database to avoid the connection
overhead. If you can't use persistent connections and you are doing a
lot of new connections to the database, you may want to change the value
of the @code{thread_cache_size} variable. @xref{Server parameters}.
@item
Always check that all your queries really use the indexes you have created
in the tables. In MySQL you can do this with the @code{EXPLAIN}
command. @xref{EXPLAIN, Explain, Explain, manual}.
@item
Try to avoid complex @code{SELECT} queries on MyISAM tables that are updated a
lot. This is to avoid problems with table locking.
@item
The new @code{MyISAM} tables can insert rows in a table without deleted
rows at the same time another table is reading from it.  If this is important
for you, you should consider methods where you don't have to delete rows
or run @code{OPTIMIZE TABLE} after you have deleted a lot of rows.
@item
Use @code{ALTER TABLE ... ORDER BY expr1,expr2...} if you mostly
retrieve rows in expr1,expr2... order.  By using this option after big
changes to the table, you may be able to get higher performance.
@item
In some cases it may make sense to introduce a column that is 'hashed'
based on information from other columns. If this column is short and
reasonably unique it may be much faster than a big index on many
columns. In MySQL it's very easy to use this extra column:
@code{SELECT * FROM table_name WHERE hash=MD5(CONCAT(col1,col2))
AND col_1='constant' AND col_2='constant'}
@item
For tables that change a lot you should try to avoid all @code{VARCHAR}
or @code{BLOB} columns. You will get dynamic row length as soon as you
are using a single @code{VARCHAR} or @code{BLOB} column. @xref{Table
types}.
@item
It's not normally useful to split a table into different tables just
because the rows gets 'big'. To access a row, the biggest performance
hit is the disk seek to find the first byte of the row. After finding
the data most new disks can read the whole row fast enough for most
applications. The only cases where it really matters to split up a table is if
it's a dynamic row size table (see above) that you can change to a fixed
row size, or if you very often need to scan the table and don't need
most of the columns. @xref{Table types}.
@item
If you very often need to calculate things based on information from a
lot of rows (like counts of things), it's probably much better to
introduce a new table and update the counter in real time. An update of
type @code{UPDATE table set count=count+1 where index_column=constant}
is very fast!

This is really important when you use databases like MySQL that
only have table locking (multiple readers / single writers). This will
also give better performance with most databases, as the row locking
manager in this case will have less to do.
@item
If you need to collect statistics from big log tables, use summary tables
instead of scanning the whole table. Maintaining the summaries should be
much faster than trying to do statistics 'live'. It's much faster to
regenerate new summary tables from the logs when things change
(depending on business decisions) than to have to change the running
application!
@item
If possible, one should classify reports as 'live' or 'statistical',
where data needed for statistical reports are only generated based on
summary tables that are generated from the actual data.
@item
Take advantage of the fact that columns have default values. Insert
values explicitly only when the value to be inserted differs from the
default. This reduces the parsing that MySQL need to do and
improves the insert speed.
@item
In some cases it's convenient to pack and store data into a blob. In this
case you have to add some extra code in your application to pack/unpack
things in the blob, but this may save a lot of accesses at some stage.
This is practical when you have data that doesn't conform to a static
table structure.
@item
Normally you should try to keep all data non-redundant (what
is called 3rd normal form in database theory), but you should not be
afraid of duplicating things or creating summary tables if you need these
to gain more speed.
@item
Stored procedures or UDF (user-defined functions) may be a good way to
get more performance.  In this case you should, however, always have a way
to do this some other (slower) way if you use some database that doesn't
support this.
@item
You can always gain something by caching queries/answers in your
application and trying to do many inserts/updates at the same time.  If
your database supports lock tables (like MySQL and Oracle),
this should help to ensure that the index cache is only flushed once
after all updates.
@item
Use @code{INSERT /*! DELAYED */} when you do not need to know when your
data is written. This speeds things up because many records can be written
with a single disk write.
@item
Use @code{INSERT /*! LOW_PRIORITY */} when you want your selects to be
more important.
@item
Use @code{SELECT /*! HIGH_PRIORITY */} to get selects that jump the
queue. That is, the select is done even if there is somebody waiting to
do a write.
@item
Use the multi-line @code{INSERT} statement to store many rows with one
SQL command (many SQL servers supports this).
@item
Use @code{LOAD DATA INFILE} to load bigger amounts of data. This is
faster than normal inserts and will be even faster when @code{myisamchk}
is integrated in @code{mysqld}.
@item
Use @code{AUTO_INCREMENT} columns to make unique values.
@item
Use @code{OPTIMIZE TABLE} once in a while to avoid fragmentation when
using a dynamic table format. @xref{OPTIMIZE TABLE, , @code{OPTIMIZE TABLE}}.

@item
Use @code{HEAP} tables to get more speed when possible. @xref{Table
types}.
@item
When using a normal web server setup, images should be stored as
files. That is, store only a file reference in the database.  The main
reason for this is that a normal web server is much better at caching
files than database contents. So it it's much easier to get a fast
system if you are using files.
@item
Use in memory tables for non-critical data that are accessed often (like
information about the last shown banner for users that don't have
cookies).
@item
Columns with identical information in different tables should be
declared identical and have identical names. Before Version 3.23 you
got slow joins otherwise.

Try to keep the names simple (use @code{name} instead of
@code{customer_name} in the customer table). To make your names portable
to other SQL servers you should keep them shorter than 18 characters.
@item
If you need really high speed, you should take a look at the low-level
interfaces for data storage that the different SQL servers support!  For
example, by accessing the MySQL @code{MyISAM} directly, you could
get a speed increase of 2-5 times compared to using the SQL interface.
To be able to do this the data must be on the same server as
the application, and usually it should only be accessed by one process
(because external file locking is really slow).  One could eliminate the
above problems by introducing low-level @code{MyISAM} commands in the
MySQL server (this could be one easy way to get more
performance if needed).  By carefully designing the database interface,
it should be quite easy to support this types of optimisation.
@item
In many cases it's faster to access data from a database (using a live
connection) than accessing a text file, just because the database is
likely to be more compact than the text file (if you are using numerical
data), and this will involve fewer disk accesses.  You will also save
code because you don't have to parse your text files to find line and
column boundaries.
@item
You can also use replication to speed things up. @xref{Replication}.
@item
Declaring a table with @code{DELAY_KEY_WRITE=1} will make the updating of
indexes faster, as these are not logged to disk until the file is closed.
The downside is that you should run @code{myisamchk} on these tables before
you start @code{mysqld} to ensure that they are okay if something killed
@code{mysqld} in the middle.  As the key information can always be generated
from the data, you should not lose anything by using @code{DELAY_KEY_WRITE}.
@end itemize


@node Locking Issues, Optimising Database Structure, Query Speed, MySQL Optimisation
@section Locking Issues


@menu
* Internal locking::            How MySQL Locks Tables
* Table locking::               Table Locking Issues
@end menu


@node Internal locking, Table locking, Locking Issues, Locking Issues
@subsection How MySQL Locks Tables

@cindex internal locking
@cindex locking, tables
@cindex tables, locking

You can find a discussion about different locking methods in the appendix.
@xref{Locking methods}.

All locking in MySQL is deadlock-free, except for @code{InnoDB} and
@code{BDB} type tables.
This is managed by always
requesting all needed locks at once at the beginning of a query and always
locking the tables in the same order.

@code{InnoDB} type tables automatically acquire their row locks and
@code{BDB} type tables
their page locks during the processing of SQL statements, not at the start
of the transaction.

The locking method MySQL uses for @code{WRITE} locks works as follows:

@itemize @bullet
@item
If there are no locks on the table, put a write lock on it.
@item
Otherwise, put the lock request in the write lock queue.
@end itemize

The locking method MySQL uses for @code{READ} locks works as follows:

@itemize @bullet
@item
If there are no write locks on the table, put a read lock on it.
@item
Otherwise, put the lock request in the read lock queue.
@end itemize

When a lock is released, the lock is made available to the threads
in the write lock queue, then to the threads in the read lock queue.

This means that if you have many updates on a table, @code{SELECT}
statements will wait until there are no more updates.

To work around this for the case where you want to do many @code{INSERT} and
@code{SELECT} operations on a table, you can insert rows in a temporary
table and update the real table with the records from the temporary table
once in a while.

This can be done with the following code:
@example
mysql> LOCK TABLES real_table WRITE, insert_table WRITE;
mysql> INSERT INTO real_table SELECT * FROM insert_table;
mysql> TRUNCATE TABLE insert_table;
mysql> UNLOCK TABLES;
@end example

You can use the @code{LOW_PRIORITY} options with @code{INSERT},
@code{UPDATE} or @code{DELETE} or @code{HIGH_PRIORITY} with
@code{SELECT} if you want to prioritise retrieval in some specific
cases.  You can also start @code{mysqld} with @code{--low-priority-updates}
to get the same behaveour.

Using @code{SQL_BUFFER_RESULT} can also help making table locks shorter.
@xref{SELECT}.

You could also change the locking code in @file{mysys/thr_lock.c} to use a
single queue.  In this case, write locks and read locks would have the same
priority, which might help some applications.


@node Table locking,  , Internal locking, Locking Issues
@subsection Table Locking Issues

@cindex problems, table locking

The table locking code in MySQL is deadlock free.

MySQL uses table locking (instead of row locking or column
locking) on all table types, except @code{InnoDB} and @code{BDB} tables,
to achieve a very
high lock speed.  For large tables, table locking is much better than
row locking for most applications, but there are, of course, some
pitfalls.

For @code{InnoDB} and @code{BDB} tables, MySQL only uses table
locking if you explicitly lock the table with @code{LOCK TABLES}.
For these table types we recommend you to not use
@code{LOCK TABLES} at all, because @code{InnoDB} uses automatic
row level locking and @code{BDB} uses page level locking to
ensure transaction isolation.

In MySQL Version 3.23.7 and above, you can insert rows into
@code{MyISAM} tables at the same time other threads are reading from the
table.  Note that currently this only works if there are no holes after
deleted rows in the table at the time the insert is made. When all holes
has been filled with new data, concurrent inserts will automatically be
enabled again.

Table locking enables many threads to read from a table at the same
time, but if a thread wants to write to a table, it must first get
exclusive access.  During the update, all other threads that want to
access this particular table will wait until the update is ready.

As updates on tables normally are considered to be more important than
@code{SELECT}, all statements that update a table have higher priority
than statements that retrieve information from a table. This should
ensure that updates are not 'starved' because one issues a lot of heavy
queries against a specific table. (You can change this by using
LOW_PRIORITY with the statement that does the update or
@code{HIGH_PRIORITY} with the @code{SELECT} statement.)

Starting from MySQL Version 3.23.7 one can use the
@code{max_write_lock_count} variable to force MySQL to
temporary give all @code{SELECT} statements, that wait for a table, a
higher priority after a specific number of inserts on a table.

Table locking is, however, not very good under the following senario:

@itemize @bullet
@item
A client issues a @code{SELECT} that takes a long time to run.
@item
Another client then issues an @code{UPDATE} on a used table. This client
will wait until the @code{SELECT} is finished.
@item
Another client issues another @code{SELECT} statement on the same table. As
@code{UPDATE} has higher priority than @code{SELECT}, this @code{SELECT}
will wait for the @code{UPDATE} to finish.  It will also wait for the first
@code{SELECT} to finish!
@item
A thread is waiting for something like @code{full disk}, in which case all
threads that wants to access the problem table will also be put in a waiting
state until more disk space is made available.
@end itemize

Some possible solutions to this problem are:

@itemize @bullet
@item
Try to get the @code{SELECT} statements to run faster. You may have to create
some summary tables to do this.

@item
Start @code{mysqld} with @code{--low-priority-updates}.  This will give
all statements that update (modify) a table lower priority than a @code{SELECT}
statement. In this case the last @code{SELECT} statement in the previous
scenario would execute before the @code{INSERT} statement.

@item
You can give a specific @code{INSERT}, @code{UPDATE}, or @code{DELETE}
statement lower priority with the @code{LOW_PRIORITY} attribute.

@item
Start @code{mysqld} with a low value for @strong{max_write_lock_count} to give
@code{READ} locks after a certain number of @code{WRITE} locks.

@item
You can specify that all updates from a specific thread should be done with
low priority by using the SQL command: @code{SET SQL_LOW_PRIORITY_UPDATES=1}.
@xref{SET OPTION, , @code{SET OPTION}}.

@item
You can specify that a specific @code{SELECT} is very important with the
@code{HIGH_PRIORITY} attribute. @xref{SELECT, , @code{SELECT}}.

@item
If you have problems with @code{INSERT} combined with @code{SELECT},
switch to use the new @code{MyISAM} tables as these support concurrent
@code{SELECT}s and @code{INSERT}s.

@item
If you mainly mix @code{INSERT} and @code{SELECT} statements, the
@code{DELAYED} attribute to @code{INSERT} will probably solve your problems.
@xref{INSERT, , @code{INSERT}}.

@item
If you have problems with @code{SELECT} and @code{DELETE}, the @code{LIMIT}
option to @code{DELETE} may help. @xref{DELETE, , @code{DELETE}}.
@end itemize


@node Optimising Database Structure, Optimising the Server, Locking Issues, MySQL Optimisation
@section Optimising Database Structure


@menu
* Design::                      Design Choices
* Data size::                   Get Your Data as Small as Possible
* MySQL indexes::               How MySQL Uses Indexes
* Indexes::                     Column Indexes
* Multiple-column indexes::     Multiple-Column Indexes
* Open tables::                 Why So Many Open tables?
* Table cache::                 How MySQL Opens and Closes Tables
* Creating many tables::        Drawbacks to Creating Large Numbers of Tables in the Same Database
@end menu


@node Design, Data size, Optimising Database Structure, Optimising Database Structure
@subsection Design Choices

@cindex design, choices
@cindex database design
@cindex storage of data

MySQL keeps row data and index data in separate files. Many (almost
all) other databases mix row and index data in the same file. We believe that
the MySQL choice is better for a very wide range of modern systems.

Another way to store the row data is to keep the information for each
column in a separate area (examples are SDBM and Focus). This will cause a
performance hit for every query that accesses more than one column. Because
this degenerates so quickly when more than one column is accessed,
we believe that this model is not good for general purpose databases.

The more common case is that the index and data are stored together
(like in Oracle/Sybase et al). In this case you will find the row
information at the leaf page of the index. The good thing with this
layout is that it, in many cases, depending on how well the index is
cached, saves a disk read.  The bad things with this layout are:

@itemize @bullet
@item
Table scanning is much slower because you have to read through the indexes
to get at the data.
@item
You can't use only the index table to retrieve data for a query.
@item
You lose a lot of space, as you must duplicate indexes from the nodes
(as you can't store the row in the nodes).
@item
Deletes will degenerate the table over time (as indexes in nodes are
usually not updated on delete).
@item
It's harder to cache only the index data.
@end itemize


@node Data size, MySQL indexes, Design, Optimising Database Structure
@subsection Get Your Data as Small as Possible

@cindex data, size
@cindex reducing, data size
@cindex storage space, minimising
@cindex tables, improving performance
@cindex performance, improving

One of the most basic optimisation is to get your data (and indexes) to
take as little space on the disk (and in memory) as possible. This can
give huge improvements because disk reads are faster and normally less
main memory will be used. Indexing also takes less resources if
done on smaller columns.

MySQL supports a lot of different table types and row formats.
Choosing the right table format may give you a big performance gain.
@xref{Table types}.

You can get better performance on a table and minimise storage space
using the techniques listed here:

@itemize @bullet
@item
Use the most efficient (smallest) types possible. MySQL has
many specialised types that save disk space and memory.

@item
Use the smaller integer types if possible to get smaller tables. For
example, @code{MEDIUMINT} is often better than @code{INT}.

@item
Declare columns to be @code{NOT NULL} if possible. It makes everything
faster and you save one bit per column. Note that if you really need
@code{NULL} in your application you should definitely use it. Just avoid
having it on all columns by default.

@item
If you don't have any variable-length columns (@code{VARCHAR},
@code{TEXT}, or @code{BLOB} columns), a fixed-size record format is
used. This is faster but unfortunately may waste some space.
@xref{MyISAM table formats}.

@item
The primary index of a table should be as short as possible. This makes
identification of one row easy and efficient.

@item
For each table, you have to decide which storage/index method to
use. @xref{Table types}.

@item
Only create the indexes that you really need. Indexes are good for
retrieval but bad when you need to store things fast. If you mostly
access a table by searching on a combination of columns, make an index
on them. The first index part should be the most used column. If you are
@strong{always} using many columns, you should use the column with more duplicates
first to get better compression of the index.

@item
If it's very likely that a column has a unique prefix on the first number
of characters, it's better to only index this prefix. MySQL
supports an index on a part of a character column. Shorter indexes are
faster not only because they take less disk space but also because they
will give you more hits in the index cache and thus fewer disk
seeks. @xref{Server parameters}.

@item
In some circumstances it can be beneficial to split into two a table that is
scanned very often. This is especially true if it is a dynamic
format table and it is possible to use a smaller static format table that
can be used to find the relevant rows when scanning the table.
@end itemize


@node MySQL indexes, Indexes, Data size, Optimising Database Structure
@subsection How MySQL Uses Indexes

@cindex indexes, use of

Indexes are used to find rows with a specific value of one column
fast. Without an index MySQL has to start with the first record
and then read through the whole table until it finds the relevant
rows. The bigger the table, the more this costs. If the table has an index
for the columns in question, MySQL can quickly get a position to
seek to in the middle of the datafile without having to look at all the
data. If a table has 1000 rows, this is at least 100 times faster than
reading sequentially. Note that if you need to access almost all 1000
rows it is faster to read sequentially because we then avoid disk seeks.

All MySQL indexes (@code{PRIMARY}, @code{UNIQUE}, and
@code{INDEX}) are stored in B-trees. Strings are automatically prefix-
and end-space compressed. @xref{CREATE INDEX, , @code{CREATE INDEX}}.

Indexes are used to:
@itemize @bullet
@item
Quickly find the rows that match a @code{WHERE} clause.

@item
Retrieve rows from other tables when performing joins.

@item
Find the @code{MAX()} or @code{MIN()} value for a specific indexed
column. This is optimised by a preprocessor that checks if you are
using @code{WHERE} key_part_# = constant on all key parts < N.  In this case
MySQL will do a single key lookup and replace the @code{MIN()}
expression with a constant.  If all expressions are replaced with
constants, the query will return at once:

@example
SELECT MIN(key_part2),MAX(key_part2) FROM table_name where key_part1=10
@end example

@item
Sort or group a table if the sorting or grouping is done on a leftmost
prefix of a usable key (for example, @code{ORDER BY
key_part_1,key_part_2 }). The key is read in reverse order if all key
parts are followed by @code{DESC}.  @xref{ORDER BY optimisation}.

@item
In some cases a query can be optimised to retrieve values without
consulting the datafile. If all used columns for some table are numeric
and form a leftmost prefix for some key, the values may be retrieved
from the index tree for greater speed:

@example
SELECT key_part3 FROM table_name WHERE key_part1=1
@end example

@end itemize

Suppose you issue the following @code{SELECT} statement:

@example
mysql> SELECT * FROM tbl_name WHERE col1=val1 AND col2=val2;
@end example

If a multiple-column index exists on @code{col1} and @code{col2}, the
appropriate rows can be fetched directly. If separate single-column
indexes exist on @code{col1} and @code{col2}, the optimiser tries to
find the most restrictive index by deciding which index will find fewer
rows and using that index to fetch the rows.

@cindex indexes, leftmost prefix of
@cindex leftmost prefix of indexes
If the table has a multiple-column index, any leftmost prefix of the
index can be used by the optimiser to find rows. For example, if you
have a three-column index on @code{(col1,col2,col3)}, you have indexed
search capabilities on @code{(col1)}, @code{(col1,col2)}, and
@code{(col1,col2,col3)}.

MySQL can't use a partial index if the columns don't form a
leftmost prefix of the index.  Suppose you have the @code{SELECT}
statements shown here:

@example
mysql> SELECT * FROM tbl_name WHERE col1=val1;
mysql> SELECT * FROM tbl_name WHERE col2=val2;
mysql> SELECT * FROM tbl_name WHERE col2=val2 AND col3=val3;
@end example

If an index exists on @code{(col1,col2,col3)}, only the first query
shown above uses the index. The second and third queries do involve
indexed columns, but @code{(col2)} and @code{(col2,col3)} are not
leftmost prefixes of @code{(col1,col2,col3)}.

@findex LIKE, and indexes
@findex LIKE, and wildcards
@cindex indexes, and @code{LIKE}
@cindex wildcards, and @code{LIKE}
MySQL also uses indexes for @code{LIKE} comparisons if the argument
to @code{LIKE} is a constant string that doesn't start with a wildcard
character.  For example, the following @code{SELECT} statements use indexes:

@example
mysql> SELECT * FROM tbl_name WHERE key_col LIKE "Patrick%";
mysql> SELECT * FROM tbl_name WHERE key_col LIKE "Pat%_ck%";
@end example

In the first statement, only rows with @code{"Patrick" <= key_col <
"Patricl"} are considered.  In the second statement, only rows with
@code{"Pat" <= key_col < "Pau"} are considered.

The following @code{SELECT} statements will not use indexes:
@example
mysql> SELECT * FROM tbl_name WHERE key_col LIKE "%Patrick%";
mysql> SELECT * FROM tbl_name WHERE key_col LIKE other_col;
@end example

In the first statement, the @code{LIKE} value begins with a wildcard
character.  In the second statement, the @code{LIKE} value is not a
constant.

MySQL 4.0 does another optimization on @code{LIKE}.  If you use
@code{... LIKE "%string%"} and @code{string} is longer than 3 characters,
MySQL will use the @code{Turbo Boyer-Moore} algorithm to initialise the
pattern for the string and then use this pattern to perform the search
quicker.

@findex IS NULL, and indexes
@cindex indexes, and @code{IS NULL}
Searching using @code{column_name IS NULL} will use indexes if column_name
is an index.

MySQL normally uses the index that finds the least number of rows. An
index is used for columns that you compare with the following operators:
@code{=}, @code{>}, @code{>=}, @code{<}, @code{<=}, @code{BETWEEN}, and a
@code{LIKE} with a non-wildcard prefix like @code{'something%'}.

Any index that doesn't span all @code{AND} levels in the @code{WHERE} clause
is not used to optimise the query. In other words:  To be able to use an
index, a prefix of the index must be used in every @code{AND} group.

The following @code{WHERE} clauses use indexes:
@example
... WHERE index_part1=1 AND index_part2=2 AND other_column=3
... WHERE index=1 OR A=10 AND index=2      /* index = 1 OR index = 2 */
... WHERE index_part1='hello' AND index_part_3=5
          /* optimised like "index_part1='hello'" */
... WHERE index1=1 and index2=2 or index1=3 and index3=3;
          /* Can use index on index1 but not on index2 or index 3 */
@end example

These @code{WHERE} clauses do @strong{NOT} use indexes:
@example
... WHERE index_part2=1 AND index_part3=2  /* index_part_1 is not used */
... WHERE index=1 OR A=10                  /* Index is not used in
                                                        both AND parts */
... WHERE index_part1=1 OR index_part2=10  /* No index spans all rows  */
@end example

Note that in some cases MySQL will not use an index, even if one
would be available.  Some of the cases where this happens are:

@itemize @bullet
@item
If the use of the index would require MySQL to access more
than 30 % of the rows in the table.  (In this case a table scan is
probably much faster, as this will require us to do much fewer seeks.)
Note that if such a query uses @code{LIMIT} to only retrieve
part of the rows, MySQL will use an index anyway, as it can
much more quickly find the few rows to return in the result.
@item
If the index range may contain @code{NULL} values and you are using
@code{ORDER BY ... DESC}
@end itemize

@node Indexes, Multiple-column indexes, MySQL indexes, Optimising Database Structure
@subsection Column Indexes

@cindex indexes, columns
@cindex columns, indexes
@cindex keys

All MySQL column types can be indexed.  Use of indexes on the
relevant columns is the best way to improve the performance of @code{SELECT}
operations.

The maximum number of keys and the maximum index length is defined per
table handler. @xref{Table types}. You can with all table handlers have
at least 16 keys and a total index length of at least 256 bytes.

For @code{CHAR} and @code{VARCHAR} columns, you can index a prefix of a
column.  This is much faster and requires less disk space than indexing the
whole column.  The syntax to use in the @code{CREATE TABLE} statement to
index a column prefix looks like this:

@example
KEY index_name (col_name(length))
@end example

The example here creates an index for the first 10 characters of the
@code{name} column:

@example
mysql> CREATE TABLE test (
    ->        name CHAR(200) NOT NULL,
    ->        KEY index_name (name(10)));
@end example

For @code{BLOB} and @code{TEXT} columns, you must index a prefix of the
column.  You cannot index the entire column.

In MySQL Version 3.23.23 or later, you can also create special
@strong{FULLTEXT} indexes. They are used for full-text search. Only the
@code{MyISAM} table type supports @code{FULLTEXT} indexes. They can be
created only from @code{VARCHAR} and @code{TEXT} columns.
Indexing always happens over the entire column and partial indexing is not
supported. See @ref{Fulltext Search} for details.

@node Multiple-column indexes, Open tables, Indexes, Optimising Database Structure
@subsection Multiple-Column Indexes

@cindex multi-column indexes
@cindex indexes, multi-column
@cindex keys, multi-column

MySQL can create indexes on multiple columns.  An index may
consist of up to 15 columns. (On @code{CHAR} and @code{VARCHAR} columns you
can also use a prefix of the column as a part of an index.)

A multiple-column index can be considered a sorted array containing values
that are created by concatenating the values of the indexed columns.

MySQL uses multiple-column indexes in such a way that queries are
fast when you specify a known quantity for the first column of the index in a
@code{WHERE} clause, even if you don't specify values for the other columns.

Suppose a table is created using the following specification:

@example
mysql> CREATE TABLE test (
    ->       id INT NOT NULL,
    ->       last_name CHAR(30) NOT NULL,
    ->       first_name CHAR(30) NOT NULL,
    ->       PRIMARY KEY (id),
    ->       INDEX name (last_name,first_name));
@end example

Then the index @code{name} is an index over @code{last_name} and
@code{first_name}.  The index will be used for queries that specify
values in a known range for @code{last_name}, or for both @code{last_name}
and @code{first_name}.
Therefore, the @code{name} index will be used in the following queries:

@example
mysql> SELECT * FROM test WHERE last_name="Widenius";

mysql> SELECT * FROM test WHERE last_name="Widenius"
    ->                    AND first_name="Michael";

mysql> SELECT * FROM test WHERE last_name="Widenius"
    ->                    AND (first_name="Michael" OR first_name="Monty");

mysql> SELECT * FROM test WHERE last_name="Widenius"
    ->                    AND first_name >="M" AND first_name < "N";
@end example

However, the @code{name} index will NOT be used in the following queries:

@example
mysql> SELECT * FROM test WHERE first_name="Michael";

mysql> SELECT * FROM test WHERE last_name="Widenius"
    ->                    OR first_name="Michael";
@end example

For more information on the manner in which MySQL uses indexes to
improve query performance, see @ref{MySQL indexes, , MySQL
indexes}.

@node Open tables, Table cache, Multiple-column indexes, Optimising Database Structure
@subsection Why So Many Open tables?

@cindex tables, open
@cindex open tables

When you run @code{mysqladmin status}, you'll see something like this:

@example
Uptime: 426 Running threads: 1 Questions: 11082 Reloads: 1 Open tables: 12
@end example

This can be somewhat perplexing if you only have 6 tables.

MySQL is multi-threaded, so it may have many queries on the same table
simultaneously. To minimise the problem with two threads having
different states on the same file, the table is opened independently by
each concurrent thread. This takes some memory but will normaly increase
performance.  Wth ISAM and MyISAM tables this also requires one extra file
descriptor for the datafile. With these table types the index file
descriptor is shared between all threads.

You can read more about this topic in the next section. @xref{Table cache}.

@node Table cache, Creating many tables, Open tables, Optimising Database Structure
@subsection How MySQL Opens and Closes Tables

@findex table_cache

@cindex tables, opening
@cindex tables, closing
@cindex opening, tables
@cindex closing, tables
@cindex table cache

@code{table_cache}, @code{max_connections}, and @code{max_tmp_tables}
affect the maximum number of files the server keeps open.  If you
increase one or both of these values, you may run up against a limit
imposed by your operating system on the per-process number of open file
descriptors.  However, you can increase the limit on many systems.
Consult your OS documentation to find out how to do this, because the
method for changing the limit varies widely from system to system.

@code{table_cache} is related to @code{max_connections}.  For example,
for 200 concurrent running connections, you should have a table cache of
at least @code{200 * n}, where @code{n} is the maximum number of tables
in a join.  You also need to reserve some extra file descriptors for
temporary tables and files.

Make sure that your operating system can handle the number of open file
descriptors implied by the @code{table_cache} setting.  If
@code{table_cache} is set too high, MySQL may run out of file
descriptors and refuse connections, fail to perform queries, and be very
unreliable.  You also have to take into account that the MyISAM table
handler needs two file descriptors for each unique open table. You can
in increase the number of file descriptors available for MySQL with
the @code{--open-files-limit=#} startup option. @xref{Not enough file
handles}.

The cache of open tables will be keept at a level of @code{table_cache}
entries (default 64; this can be changed with the @code{-O
table_cache=#} option to @code{mysqld}).  Note that in MySQL may
temporarly open even more tables to be able to execute queries.

A not used table is closed and removed from the table cache under the
following circumstances:

@itemize @bullet
@item
When the cache is full and a thread tries to open a table that is not in
the cache.
@item
When the cache contains more than @code{table_cache} entries and
a thread is no longer using a table.
@item
When someone executes @code{mysqladmin refresh} or
@code{mysqladmin flush-tables}.
@item
When someone executes 'FLUSH TABLES'
@end itemize

When the table cache fills up, the server uses the following procedure
to locate a cache entry to use:

@itemize @bullet
@item
Tables that are not currently in use are released, in least-recently-used
order.

@item
If the cache is full and no tables can be released, but a new table needs to
be opened, the cache is temporarily extended as necessary.

@item
If the cache is in a temporarily-extended state and a table goes from in-use
to not-in-use state, the table is closed and released from the cache.
@end itemize

A table is opened for each concurrent access. This means that
if you have two threads accessing the same table or access the table
twice in the same query (with @code{AS}) the table needs to be opened twice.
The first open of any table takes two file descriptors; each additional
use of the table takes only one file descriptor.  The extra descriptor
for the first open is used for the index file; this descriptor is shared
among all threads.

If you are opening a table with the @code{HANDLER table_name OPEN}
statement, a dedicated table object is allocated for the thread.
This table object is not shared by other threads an will not be closed
until the thread calls @code{HANDLER table_name CLOSE} or the thread dies.
@xref{HANDLER, , @code{HANDLER}}.  When this happens, the table is put
back in the table_cache (if it isn't full).

You can check if your table cache is too small by checking the mysqld
variable @code{Opened_tables}.  If this is quite big, even if you
haven't done a lot of @code{FLUSH TABLES}, you should increase your table
cache. @xref{SHOW STATUS, , @code{SHOW STATUS}}.


@node Creating many tables,  , Table cache, Optimising Database Structure
@subsection Drawbacks to Creating Large Numbers of Tables in the Same Database

@cindex tables, too many

If you have many files in a directory, open, close, and create operations will
be slow.  If you execute @code{SELECT} statements on many different tables,
there will be a little overhead when the table cache is full, because for
every table that has to be opened, another must be closed. You can reduce
this overhead by making the table cache larger.


@node Optimising the Server, Disk issues, Optimising Database Structure, MySQL Optimisation
@section Optimising the MySQL Server


@menu
* System::                      System/Compile Time and Startup Parameter Tuning
* Server parameters::           Tuning Server Parameters
* Compile and link options::    How Compiling and Linking Affects the Speed of MySQL
* Memory use::                  How MySQL Uses Memory
* DNS::                         How MySQL uses DNS
* SET OPTION::                  @code{SET} Syntax
@end menu


@node System, Server parameters, Optimising the Server, Optimising the Server
@subsection System/Compile Time and Startup Parameter Tuning

@cindex compiling, optimising
@cindex system optimisation
@cindex startup parameters, tuning

We start with the system level things since some of these decisions have
to be made very early. In other cases a fast look at this part may
suffice because it not that important for the big gains. However, it is always
nice to have a feeling about how much one could gain by changing things
at this level.

The default OS to use is really important!  To get the most use of
multiple-CPU machines one should use Solaris (because the threads works
really nice) or Linux (because the 2.2 kernel has really good SMP
support). Also on 32-bit machines Linux has a 2G file-size limit by
default. Hopefully this will be fixed soon when new filesystems are
released (XFS/Reiserfs).  If you have a desperate need for files bigger
than 2G on Linux-intel 32 bit, you should get the LFS patch for the ext2
filesystem.

Because we have not run MySQL in production on that many platforms, we
advice you to test your intended platform before choosing it, if possible.

@cindex locking
Other tips:
@itemize @bullet
@item
If you have enough RAM, you could remove all swap devices. Some
operating systems will use a swap device in some contexts even if you
have free memory.
@item
Use the @code{--skip-locking} MySQL option to avoid external
locking. Note that this will not impact MySQL's functionality as
long as you only run one server. Just remember to take down the server (or
lock relevant parts) before you run @code{myisamchk}. On some system
this switch is mandatory because the external locking does not work in any
case.

The @code{--skip-locking} option is on by default when compiling with
MIT-pthreads, because @code{flock()} isn't fully supported by
MIT-pthreads on all platforms.  It's also on default for Linux
as Linux file locking are not yet safe.

The only case when you can't use @code{--skip-locking} is if you run
multiple MySQL @emph{servers} (not clients) on the same data,
or run @code{myisamchk} on the table without first flushing and locking
the @code{mysqld} server tables first.

You can still use @code{LOCK TABLES}/@code{UNLOCK TABLES} even if you
are using @code{--skip-locking}
@end itemize


@node Server parameters, Compile and link options, System, Optimising the Server
@subsection Tuning Server Parameters

@cindex parameters, server
@cindex @code{mysqld} server, buffer sizes
@cindex buffer sizes, @code{mysqld} server
@cindex startup parameters

You can get the default buffer sizes used by the @code{mysqld} server
with this command:

@example
shell> mysqld --help
@end example

@cindex @code{mysqld} options
@cindex variables, @code{mysqld}
This command produces a list of all @code{mysqld} options and configurable
variables. The output includes the default values and looks something
like this:

@example
Possible variables for option --set-variable (-O) are:
back_log                 current value: 5
bdb_cache_size           current value: 1048540
binlog_cache_size        current value: 32768
connect_timeout          current value: 5
delayed_insert_timeout   current value: 300
delayed_insert_limit     current value: 100
delayed_queue_size       current value: 1000
flush_time               current value: 0
interactive_timeout      current value: 28800
join_buffer_size         current value: 131072
key_buffer_size          current value: 1048540
lower_case_table_names   current value: 0
long_query_time          current value: 10
max_allowed_packet       current value: 1048576
max_binlog_cache_size    current value: 4294967295
max_connections          current value: 100
max_connect_errors       current value: 10
max_delayed_threads      current value: 20
max_heap_table_size      current value: 16777216
max_join_size            current value: 4294967295
max_sort_length          current value: 1024
max_tmp_tables           current value: 32
max_write_lock_count     current value: 4294967295
myisam_sort_buffer_size  current value: 8388608
net_buffer_length        current value: 16384
net_retry_count          current value: 10
net_read_timeout         current value: 30
net_write_timeout        current value: 60
query_buffer_size        current value: 0
record_buffer            current value: 131072
record_rnd_buffer        current value: 131072
slow_launch_time         current value: 2
sort_buffer              current value: 2097116
table_cache              current value: 64
thread_concurrency       current value: 10
tmp_table_size           current value: 1048576
thread_stack             current value: 131072
wait_timeout             current value: 28800
@end example

If there is a @code{mysqld} server currently running, you can see what
values it actually is using for the variables by executing this command:

@example
shell> mysqladmin variables
@end example

You can find a full description for all variables in the @code{SHOW VARIABLES}
section in this manual. @xref{SHOW VARIABLES}.

You can also see some statistics from a running server by issuing the command
@code{SHOW STATUS}. @xref{SHOW STATUS}.

MySQL uses algorithms that are very scalable, so you can usually
run with very little memory. If you, however, give MySQL more
memory, you will normally also get better performance.

When tuning a MySQL server, the two most important variables to use
are @code{key_buffer_size} and @code{table_cache}.  You should first feel
confident that you have these right before trying to change any of the
other variables.

If you have much memory (>=256M) and many tables and want maximum performance
with a moderate number of clients, you should use something like this:

@example
shell> safe_mysqld -O key_buffer=64M -O table_cache=256 \
           -O sort_buffer=4M -O record_buffer=1M &
@end example

If you have only 128M and only a few tables, but you still do a lot of
sorting, you can use something like:

@example
shell> safe_mysqld -O key_buffer=16M -O sort_buffer=1M
@end example

If you have little memory and lots of connections, use something like this:

@example
shell> safe_mysqld -O key_buffer=512k -O sort_buffer=100k \
           -O record_buffer=100k &
@end example

or even:

@example
shell> safe_mysqld -O key_buffer=512k -O sort_buffer=16k \
           -O table_cache=32 -O record_buffer=8k -O net_buffer=1K &
@end example

If you are doing a @code{GROUP BY} or @code{ORDER BY} on files that are
much bigger than your available memory you should increase the value of
@code{record_rnd_buffer} to speed up the reading of rows after the sorting
is done.

When you have installed MySQL, the @file{support-files} directory will
contain some different @file{my.cnf} example files, @file{my-huge.cnf},
@file{my-large.cnf}, @file{my-medium.cnf}, and @file{my-small.cnf}, you can
use as a base to optimise your system.

If there are very many connections, ``swapping problems'' may occur unless
@code{mysqld} has been configured to use very little memory for each
connection. @code{mysqld} performs better if you have enough memory for all
connections, of course.

Note that if you change an option to @code{mysqld}, it remains in effect only
for that instance of the server.

To see the effects of a parameter change, do something like this:

@example
shell> mysqld -O key_buffer=32m --help
@end example

Make sure that the @code{--help} option is last; otherwise, the effect of any
options listed after it on the command-line will not be reflected in the
output.


@node Compile and link options, Memory use, Server parameters, Optimising the Server
@subsection How Compiling and Linking Affects the Speed of MySQL

@cindex linking, speed
@cindex compiling, speed
@cindex speed, compiling
@cindex speed, linking

Most of the following tests are done on Linux with the
MySQL benchmarks, but they should give some indication for
other operating systems and workloads.

You get the fastest executable when you link with @code{-static}.

On Linux, you will get the fastest code when compiling with @code{pgcc}
and @code{-O3}. To compile @file{sql_yacc.cc} with these options, you
need about 200M memory because @code{gcc/pgcc} needs a lot of memory to
make all functions inline. You should also set @code{CXX=gcc} when
configuring MySQL to avoid inclusion of the @code{libstdc++}
library (it is not needed). Note that with some versions of @code{pgcc},
the resulting code will only run on true Pentium processors, even if you
use the compiler option that you want the resulting code to be working on
all x586 type processors (like AMD).

By just using a better compiler and/or better compiler options you can
get a 10-30 % speed increase in your application.  This is particularly
important if you compile the SQL server yourself!

We have tested both the Cygnus CodeFusion and Fujitsu compilers, but
when we tested them, neither was sufficiently bug free to allow
MySQL to be compiled with optimisations on.

When you compile MySQL you should only include support for the
character sets that you are going to use. (Option @code{--with-charset=xxx}.)
The standard MySQL binary distributions are compiled with support
for all character sets.

Here is a list of some measurements that we have done:
@itemize @bullet
@item
If you use @code{pgcc} and compile everything with @code{-O6}, the
@code{mysqld} server is 1% faster than with @code{gcc} 2.95.2.

@item
If you link dynamically (without @code{-static}), the result is 13%
slower on Linux.  Note that you still can use a dynamic linked
MySQL library. It is only the server that is critical for
performance.

@item
If you strip your @code{mysqld} binary with @code{strip libexec/mysqld},
the resulting binary can be up to 4 % faster.

@item
If you connect using TCP/IP rather than Unix sockets, the result is 7.5%
slower on the same computer. (If you are connection to @code{localhost},
MySQL will, by default, use sockets.)

@item
If you connect using TCP/IP from another computer over a 100M Ethernet,
things will be 8-11 % slower.

@item
If you compile with @code{--with-debug=full}, then you will loose 20 %
for most queries, but some queries may take substantially longer (The
MySQL benchmarks ran 35 % slower)
If you use @code{--with-debug}, then you will only loose 15 %.
By starting a @code{mysqld} version compiled with @code{--with-debug=full}
with @code{--skip-safemalloc} the end result should be close to when
configuring with @code{--with-debug}.

@item
On a Sun SPARCstation 20, SunPro C++ 4.2 is 5 % faster than @code{gcc} 2.95.2.

@item
Compiling with @code{gcc} 2.95.2 for UltraSPARC with the option
@code{-mcpu=v8 -Wa,-xarch=v8plusa} gives 4 % more performance.

@item
On Solaris 2.5.1, MIT-pthreads is 8-12% slower than Solaris native
threads on a single processor. With more load/CPUs the difference should
get bigger.

@item
Running with @code{--log-bin} makes @strong{[MySQL} 1 % slower.

@item
Compiling on Linux-x86 using gcc without frame pointers
@code{-fomit-frame-pointer} or @code{-fomit-frame-pointer -ffixed-ebp}
makes @code{mysqld} 1-4% faster.
@end itemize

The MySQL-Linux distribution provided by MySQL AB used
to be compiled with @code{pgcc}, but we had to go back to regular gcc
because of a bug in @code{pgcc} that would generate the code that does
not run on AMD. We will continue using gcc until that bug is resolved.
In the meantime, if you have a non-AMD machine, you can get a faster
binary by compiling with @code{pgcc}.  The standard MySQL
Linux binary is linked statically to get it faster and more portable.


@node Memory use, DNS, Compile and link options, Optimising the Server
@subsection How MySQL Uses Memory

@cindex memory use

The following list indicates some of the ways that the @code{mysqld} server
uses memory.  Where applicable, the name of the server variable relevant
to the memory use is given:

@itemize @bullet
@item
The key buffer (variable @code{key_buffer_size}) is shared by all
threads; other buffers used by the server are allocated as
needed. @xref{Server parameters}.

@item
Each connection uses some thread-specific space: A stack (default 64K,
variable @code{thread_stack}), a connection buffer (variable
@code{net_buffer_length}), and a result buffer (variable
@code{net_buffer_length}). The connection buffer and result buffer are
dynamically enlarged up to @code{max_allowed_packet} when needed.  When
a query is running, a copy of the current query string is also allocated.

@item
All threads share the same base memory.

@item
Only the compressed ISAM / MyISAM tables are memory mapped.  This is
because the 32-bit memory space of 4GB is not large enough for most
big tables. When systems with a 64-bit address space become more
common we may add general support for memory mapping.

@item
Each request doing a sequential scan over a table allocates a read buffer
(variable @code{record_buffer}).

@item
When reading rows in 'random' order (for example after a sort) a
random-read buffer is allocated to avoid disk seeks.
(variable @code{record_rnd_buffer}).

@item
All joins are done in one pass, and most joins can be done without even
using a temporary table. Most temporary tables are memory-based (HEAP)
tables.  Temporary tables with a big record length (calculated as the
sum of all column lengths) or that contain @code{BLOB} columns are
stored on disk.

One problem in MySQL versions before Version 3.23.2 is that if a HEAP table
exceeds the size of @code{tmp_table_size}, you get the error @code{The
table tbl_name is full}. In newer versions this is handled by
automatically changing the in-memory (HEAP) table to a disk-based
(MyISAM) table as necessary. To work around this problem, you can
increase the temporary table size by setting the @code{tmp_table_size}
option to @code{mysqld}, or by setting the SQL option
@code{SQL_BIG_TABLES} in the client program.  @xref{SET OPTION, ,
@code{SET OPTION}}. In MySQL Version 3.20, the maximum size of the
temporary table was @code{record_buffer*16}, so if you are using this
version, you have to increase the value of @code{record_buffer}. You can
also start @code{mysqld} with the @code{--big-tables} option to always
store temporary tables on disk. However, this will affect the speed of
many complicated queries.

@item
Most requests doing a sort allocates a sort buffer and 0-2 temporary
files depending on the result set size. @xref{Temporary files}.

@item
Almost all parsing and calculating is done in a local memory store. No
memory overhead is needed for small items and the normal slow memory
allocation and freeing is avoided. Memory is allocated only for
unexpectedly large strings (this is done with @code{malloc()} and
@code{free()}).

@item
Each index file is opened once and the datafile is opened once for each
concurrently running thread. For each concurrent thread, a table structure,
column structures for each column, and a buffer of size @code{3 * n} is
allocated (where @code{n} is the maximum row length, not counting @code{BLOB}
columns). A @code{BLOB} uses 5 to 8 bytes plus the length of the @code{BLOB}
data. The @code{ISAM}/@code{MyISAM} table handlers will use one extra row
buffer for internal usage.

@item
For each table having @code{BLOB} columns, a buffer is enlarged dynamically
to read in larger @code{BLOB} values. If you scan a table, a buffer as large
as the largest @code{BLOB} value is allocated.

@item
Table handlers for all in-use tables are saved in a cache and managed as a
FIFO. Normally the cache has 64 entries. If a table has been used by two
running threads at the same time, the cache contains two entries for the
table. @xref{Table cache}.

@item
A @code{mysqladmin flush-tables} command closes all tables that are not in
use and marks all in-use tables to be closed when the currently executing
thread finishes. This will effectively free most in-use memory.
@end itemize

@code{ps} and other system status programs may report that @code{mysqld}
uses a lot of memory. This may be caused by thread-stacks on different
memory addresses. For example, the Solaris version of @code{ps} counts
the unused memory between stacks as used memory. You can verify this by
checking available swap with @code{swap -s}. We have tested
@code{mysqld} with commercial memory-leakage detectors, so there should
be no memory leaks.


@node DNS, SET OPTION, Memory use, Optimising the Server
@subsection How MySQL uses DNS

@cindex DNS
@cindex hostname caching

When a new thread connects to @code{mysqld}, @code{mysqld} will span a
new thread to handle the request.  This thread will first check if the
hostname is in the hostname cache. If not the thread will call
@code{gethostbyaddr_r()} and @code{gethostbyname_r()} to resolve the
hostname.

If the operating system doesn't support the above thread-safe calls, the
thread will lock a mutex and call @code{gethostbyaddr()} and
@code{gethostbyname()} instead. Note that in this case no other thread
can resolve other hostnames that is not in the hostname cache until the
first thread is ready.

You can disable DNS host lookup by starting @code{mysqld} with
@code{--skip-name-resolve}.  In this case you can however only use IP
names in the MySQL privilege tables.

If you have a very slow DNS and many hosts, you can get more performance by
either disabling DNS lookop with @code{--skip-name-resolve} or by
increasing the @code{HOST_CACHE_SIZE} define (default: 128) and recompile
@code{mysqld}.

You can disable the hostname cache with @code{--skip-host-cache}. You
can clear the hostname cache with @code{FLUSH HOSTS} or @code{mysqladmin
flush-hosts}.

If you don't want to allow connections over @code{TCP/IP}, you can do this
by starting @code{mysqld} with @code{--skip-networking}.


@node SET OPTION,  , DNS, Optimising the Server
@subsection @code{SET} Syntax

@findex SET OPTION

@example
SET [OPTION] SQL_VALUE_OPTION= value, ...
@end example

@code{SET OPTION} sets various options that affect the operation of the
server or your client.  Any option you set remains in effect until the
current session ends, or until you set the option to a different value.

@table @code
@item CHARACTER SET character_set_name | DEFAULT
This maps all strings from and to the client with the given mapping.
Currently the only option for @code{character_set_name} is
@code{cp1251_koi8}, but you can easily add new mappings by editing the
@file{sql/convert.cc} file in the MySQL source distribution.  The
default mapping can be restored by using a @code{character_set_name} value of
@code{DEFAULT}.

Note that the syntax for setting the @code{CHARACTER SET} option differs
from the syntax for setting the other options.

@item PASSWORD = PASSWORD('some password')
@cindex passwords, setting
Set the password for the current user. Any non-anonymous user can change his
own password!

@item PASSWORD FOR user = PASSWORD('some password')
Set the password for a specific user on the current server host. Only a user
with access to the @code{mysql} database can do this.  The user should be
given in @code{user@@hostname} format, where @code{user} and @code{hostname}
are exactly as they are listed in the @code{User} and @code{Host} columns of
the @code{mysql.user} table entry.  For example, if you had an entry with
@code{User} and @code{Host} fields of @code{'bob'} and @code{'%.loc.gov'},
you would write:

@example
mysql> SET PASSWORD FOR bob@@"%.loc.gov" = PASSWORD("newpass");

or

mysql> UPDATE mysql.user SET password=PASSWORD("newpass")
    ->                   WHERE user="bob' AND host="%.loc.gov";
@end example

@item SQL_AUTO_IS_NULL = 0 | 1
If set to @code{1} (default) then one can find the last inserted row
for a table with an @code{AUTO_INCREMENT} column with the following construct:
@code{WHERE auto_increment_column IS NULL}.  This is used by some
ODBC programs like Access.

@item AUTOCOMMIT= 0 | 1
If set to @code{1} all changes to a table will be done at once. To start
a multi-command transaction, you have to use the @code{BEGIN}
statement. @xref{COMMIT}. If set to @code{0} you have to use @code{COMMIT} /
@code{ROLLBACK} to accept/revoke that transaction. @xref{COMMIT}.  Note
that when you change from not @code{AUTOCOMMIT} mode to
@code{AUTOCOMMIT} mode, MySQL will do an automatic
@code{COMMIT} on any open transactions.

@item SQL_BIG_TABLES = 0 | 1
@cindex table is full
If set to @code{1}, all temporary tables are stored on disk rather than in
memory.  This will be a little slower, but you will not get the error
@code{The table tbl_name is full} for big @code{SELECT} operations that
require a large temporary table.  The default value for a new connection is
@code{0} (that is, use in-memory temporary tables).

@item SQL_BIG_SELECTS = 0 | 1
If set to @code{0}, MySQL will abort if a @code{SELECT} is attempted
that probably will take a very long time. This is useful when an inadvisable
@code{WHERE} statement has been issued. A big query is defined as a
@code{SELECT} that probably will have to examine more than
@code{max_join_size} rows.  The default value for a new connection is
@code{1} (which will allow all @code{SELECT} statements).

@item SQL_BUFFER_RESULT = 0 | 1
@code{SQL_BUFFER_RESULT} will force the result from @code{SELECT}'s
to be put into a temporary table. This will help MySQL free the
table locks early and will help in cases where it takes a long time to
send the result set to the client.

@item SQL_LOW_PRIORITY_UPDATES = 0 | 1
If set to @code{1}, all @code{INSERT}, @code{UPDATE}, @code{DELETE}, and
and @code{LOCK TABLE WRITE} statements wait until there is no pending
@code{SELECT} or @code{LOCK TABLE READ} on the affected table.

@item SQL_MAX_JOIN_SIZE = value | DEFAULT
Don't allow @code{SELECT}s that will probably need to examine more than
@code{value} row combinations.  By setting this value, you can catch
@code{SELECT}s where keys are not used properly and that would probably
take a long time. Setting this to a value other than @code{DEFAULT} will reset
the @code{SQL_BIG_SELECTS} flag.  If you set the @code{SQL_BIG_SELECTS}
flag again, the @code{SQL_MAX_JOIN_SIZE} variable will be ignored.
You can set a default value for this variable by starting @code{mysqld} with
@code{-O max_join_size=#}.

@item SQL_QUERY_CACHE_TYPE = OFF | ON | DEMAND
@item SQL_QUERY_CACHE_TYPE = 0   | 1  | 2
Set query cache setting for this thread.
@multitable @columnfractions .20 .70
@item @strong{Option} @tab @strong{Description}
@item 0 or OFF @tab Don't cache or retrieve results.
@item 1 or ON @tab Cache all results except @code{SELECT SQL_NO_CACHE ...} queries.
@item 2 or DEMAND @tab Cache only @code{SELECT SQL_CACHE ...} queries.
@end multitable

@item SQL_SAFE_UPDATES = 0 | 1
If set to @code{1}, MySQL will abort if an @code{UPDATE} or
@code{DELETE} is attempted that doesn't use a key or @code{LIMIT} in the
@code{WHERE} clause. This makes it possible to catch wrong updates
when creating SQL commands by hand.

@item SQL_SELECT_LIMIT = value | DEFAULT
The maximum number of records to return from @code{SELECT} statements.  If
a @code{SELECT} has a @code{LIMIT} clause, the @code{LIMIT} takes precedence
over the value of @code{SQL_SELECT_LIMIT}.  The default value for a new
connection is ``unlimited.'' If you have changed the limit, the default value
can be restored by using a @code{SQL_SELECT_LIMIT} value of @code{DEFAULT}.

@item SQL_LOG_OFF = 0 | 1
If set to @code{1}, no logging will be done to the standard log for this
client, if the client has the @strong{process} privilege.  This does not
affect the update log!

@item SQL_LOG_UPDATE = 0 | 1
If set to @code{0}, no logging will be done to the update log for the client,
if the client has the @strong{process} privilege.  This does not affect the
standard log!

@item SQL_QUOTE_SHOW_CREATE = 0 | 1
If set to @code{1}, @code{SHOW CREATE TABLE} will quote
table and column names. This is @strong{on} by default,
for replication of tables with fancy column names to work.
@ref{SHOW CREATE TABLE, , @code{SHOW CREATE TABLE}}.

@item TIMESTAMP = timestamp_value | DEFAULT
Set the time for this client.  This is used to get the original timestamp if
you use the update log to restore rows.  @code{timestamp_value} should be a
Unix epoch timestamp, not a MySQL timestamp.

@item LAST_INSERT_ID = #
Set the value to be returned from @code{LAST_INSERT_ID()}. This is stored in
the update log when you use @code{LAST_INSERT_ID()} in a command that updates
a table.

@item INSERT_ID = #
Set the value to be used by the following @code{INSERT} or @code{ALTER TABLE}
command when inserting an @code{AUTO_INCREMENT} value.  This is mainly used
with the update log.
@end table

@xref{SET TRANSACTION}.

@node Disk issues,  , Optimising the Server, MySQL Optimisation
@section Disk Issues

@cindex disk issues
@cindex performance, disk issues

@itemize @bullet
@item
As mentioned before, disks seeks are a big performance bottleneck. This
problems gets more and more apparent when the data starts to grow so
large that effective caching becomes impossible. For large databases,
where you access data more or less randomly, you can be sure that you
will need at least one disk seek to read and a couple of disk seeks to
write things. To minimise this problem, use disks with low seek times.

@item
Increase the number of available disk spindles (and thereby reduce
the seek overhead) by either symlink files to different disks or striping
the disks.

@table @strong
@item Using symbolic links
This means that you symlink the index and/or datafile(s) from the
normal data directory to another disk (that may also be striped). This
makes both the seek and read times better (if the disks are not used for
other things). @xref{Symbolic links}.

@cindex striping, defined
@item Striping
Striping means that you have many disks and put the first block on the
first disk, the second block on the second disk, and the Nth on the
(N mod number_of_disks) disk, and so on. This means if your normal data
size is less than the stripe size (or perfectly aligned) you will get
much better performance. Note that striping is very dependent on the OS
and stripe-size. So benchmark your application with different
stripe-sizes. @xref{Custom Benchmarks}.

Note that the speed difference for striping is @strong{very} dependent
on the parameters. Depending on how you set the striping parameters and
number of disks you may get a difference in orders of magnitude. Note that
you have to choose to optimise for random or sequential access.
@end table

@item
For reliability you may want to use RAID 0+1 (striping + mirroring), but
in this case you will need 2*N drives to hold N drives of data. This is
probably the best option if you have the money for it!  You may, however,
also have to invest in some volume-management software to handle it
efficiently.

@item
A good option is to have semi-important data (that can be regenerated)
on RAID 0 disk while storing really important data (like host information
and logs) on a RAID 0+1 or RAID N disk. RAID N can be a problem if you
have many writes because of the time to update the parity bits.

@item
You may also set the parameters for the filesystem that the database
uses. One easy change is to mount the filesystem with the noatime
option. That makes it skip the updating of the last access time in the
inode and by this will avoid some disk seeks.

@item
On Linux, you can get much more performance (up to 100 % under load is
not uncommon) by using hdpram to configure your disk's interface!  The
following should be quite good hdparm options for MySQL (and
probably many other applications):

@example
hdparm -m 16 -d 1
@end example

Note that the performance/reliability when using the above depends on
your hardware, so we strongly suggest that you test your system
thoroughly after using @code{hdparm}!  Please consult the @code{hdparm}
man page for more information!  If @code{hdparm} is not used wisely,
filesystem corruption may result.  Backup everything before experimenting!

@item
On many operating systems you can mount the disks with the 'async' flag to
set the filesystem to be updated asynchronously.  If your computer is
reasonable stable, this should give you more performance without sacrificing
too much reliability.  (This flag is on by default on Linux.)

@item
If you don't need to know when a file was last accessed (which is not
really useful on a database server), you can mount your filesystems
with the noatime flag.
@end itemize

@menu
* Symbolic links::              Using Symbolic Links
@end menu


@node Symbolic links,  , Disk issues, Disk issues
@subsection Using Symbolic Links

@cindex symbolic links
@cindex links, symbolic

You can move tables and databases from the database directory to other
locations and replace them with symbolic links to the new locations.
You might want to do this, for example, to move a database to a file
system with more free space or increase the speed of your system by
spreading your tables to different disk.

The recommended way to do this, is to just symlink databases to a
different disk and only symlink tables as a last resort.

@cindex databases, symbolic links
@menu
* Symbolic links to databases::  Using Symbolic Links for Databases
* Symbolic links to tables::    Using Symbolic Links for Tables
@end menu


@node Symbolic links to databases, Symbolic links to tables, Symbolic links, Symbolic links
@subsubsection Using Symbolic Links for Databases

The way to symlink a database is to first create a directory on some
disk where you have free space and then create a symlink to it from
the MySQL database directory.

@example
shell> mkdir /dr1/databases/test
shell> ln -s /dr1/databases/test mysqld-datadir
@end example

MySQL doesn't support that you link one directory to multiple
databases.  Replacing a database directory with a symbolic link will
work fine as long as you don't make a symbolic link between databases.
Suppose you have a database @code{db1} under the MySQL data
directory, and then make a symlink @code{db2} that points to @code{db1}:

@example
shell> cd /path/to/datadir
shell> ln -s db1 db2
@end example

Now, for any table @code{tbl_a} in @code{db1}, there also appears to be
a table @code{tbl_a} in @code{db2}.  If one thread updates @code{db1.tbl_a}
and another thread updates @code{db2.tbl_a}, there will be problems.

If you really need this, you must change the following code in
@file{mysys/mf_format.c}:

@example
if (flag & 32 || (!lstat(to,&stat_buff) && S_ISLNK(stat_buff.st_mode)))
@end example

to

@example
if (1)
@end example

On Windows you can use internal symbolic links to directories by compiling
MySQL with @code{-DUSE_SYMDIR}. This allows you to put different
databases on different disks. @xref{Windows symbolic links}.


@node Symbolic links to tables,  , Symbolic links to databases, Symbolic links
@subsubsection Using Symbolic Links for Tables

@cindex databases, symbolic links

Before MySQL 4.0 you should not symlink tables, if you are not
very careful with them. The problem is that if you run @code{ALTER
TABLE}, @code{REPAIR TABLE} or @code{OPTIMIZE TABLE} on a symlinked
table, the symlinks will be removed and replaced by the original
files. This happens because the above command works by creating a
temporary file in the database directory and when the command is
complete, replace the original file with the temporary file.

You should not symlink tables on systems that don't have a fully
working @code{realpath()} call. (At least Linux and Solaris support
@code{realpath()})

In MySQL 4.0 symlinks are only fully supported for @code{MyISAM}
tables.  For other table types you will probably get strange problems
when doing any of the above mentioned commands.

The handling of symbolic links in MySQL 4.0 works the following
way (this is mostly relevant only for @code{MyISAM} tables).

@itemize @bullet
@item
In the data directory you will always have the table definition file
and the data/index files.

@item
You can symlink the index file and the datafile to different directories
independent of the other.

@item
The symlinking can be done from the operating system (if @code{mysqld} is
not running) or with the @code{INDEX/DATA DIRECTORY="path-to-dir"} command
in @code{CREATE TABLE}. @xref{CREATE TABLE}.

@item
@code{myisamchk} will not replace a symlink with the index/file but
work directly on the files the symlinks points to. Any temporary files
will be created in the same directory where the data/index file is.

@item
When you drop a table that is using symlinks, both the symlink and the
file the symlink points to is dropped.  This is a good reason to why you
should @strong{not} run @code{mysqld} as root and not allow persons to have write
access to the MySQL database directories.

@item
If you rename a table with @code{ALTER TABLE RENAME} and you don't change
database, the symlink in the database directory will be renamed to the new
name and the data/index file will be renamed accordingly.

@item
If you use @code{ALTER TABLE RENAME} to move a table to another database,
then the table will be moved to the other database directory and the old
symlinks and the files they pointed to will be deleted.

@item
If you are not using symlinks you should use the @code{--skip-symlink}
option to @code{mysqld} to ensure that no one can drop or rename a file
outside of the @code{mysqld} data directory.
@end itemize

Things that are not yet supported:

@cindex TODO, symlinks
@itemize @bullet
@item
@code{ALTER TABLE} ignores all @code{INDEX/DATA DIRECTORY="path"} options.
@item
@code{CREATE TABLE} doesn't report if the table has symbolic links.
@item
@code{mysqldump} doesn't include the symbolic links information in the output.
@item
@code{BACKUP TABLE} and @code{RESTORE TABLE} don't respect symbolic links.
@end itemize


@node Reference, Table types, MySQL Optimisation, Top
@chapter MySQL Language Reference

MySQL has a very complex, but intuitive and easy to learn SQL
interface.  This chapter describes the various commands, types, and functions
you will need to know in order to use MySQL efficiently and
effectively.  This chapter also serves as a reference to all functionality
included in MySQL.  In order to use this chapter effectively, you
may find it useful to refer to the various indexes.

@menu
* Language Structure::          Language Structure
* Column types::                Column types
* Functions::                   Functions
* Data Manipulation::           Data Manipulation: @code{SELECT}, @code{INSERT}, @code{UPDATE}, @code{DELETE}
* Data Definition::             Data Definition: @code{CREATE}, @code{DROP}, @code{ALTER}
* Basic User Commands::         Basic MySQL User Utility Commands
* Transactional Commands::      MySQL Transactional and Locking Commands
* Fulltext Search::             MySQL Full-text Search
* Query Cache::                 MySQL Query Cache
@end menu


@node Language Structure, Column types, Reference, Reference
@section Language Structure

@menu
* Literals::                    Literals: How to Write Strings and Numbers
* Legal names::                 Database, Table, Index, Column, and Alias Names
* Name case sensitivity::       Case Sensitivity in Names
* Variables::                   User Variables
* Comments::                    Comment Syntax
* Reserved words::              Is MySQL Picky About Reserved Words?
@end menu


@node Literals, Legal names, Language Structure, Language Structure
@subsection Literals: How to Write Strings and Numbers

@cindex strings, defined
@cindex strings, escaping characters
@cindex literals
@cindex escape characters
@cindex backslash, escape character

@menu
* String syntax::               Strings
* Number syntax::               Numbers
* Hexadecimal values::          Hexadecimal values
* NULL values::                 @code{NULL} values
@end menu


This section describes the various ways to write strings and numbers in
MySQL.  It also covers the various nuances and ``gotchas'' that
you may run into when dealing with these basic types in MySQL.


@node String syntax, Number syntax, Literals, Literals
@subsubsection Strings

A string is a sequence of characters, surrounded by either single quote
(@samp{'}) or double quote (@samp{"}) characters (only the single quote
if you run in ANSI mode).  Examples:

@example
'a string'
"another string"
@end example

Within a string, certain sequences have special meaning.  Each of these
sequences begins with a backslash (@samp{\}), known as the @emph{escape
character}.  MySQL recognises the following escape sequences:

@c these aren't really functions, but that's probably the most reasonable index
@table @code
@findex \0 (ASCII 0)
@findex NUL
@item \0
An ASCII 0 (@code{NUL}) character.

@findex \' (single quote)
@findex single quote (\')
@item \'
A single quote (@samp{'}) character.

@findex \" (double quote)
@findex double quote (\")
@item \"
A double quote (@samp{"}) character.

@findex \b (backspace)
@findex backspace (\b)
@item \b
A backspace character.

@findex \n (newline)
@findex newline (\n)
@item \n
A newline character.

@findex \r (carriage return)
@findex return (\r)
@findex carriage return (\r)
@item \r
A carriage return character.

@findex \t (tab)
@findex tab (\t)
@item \t
A tab character.

@findex \z (Control-Z) ASCII(26)
@findex (Control-Z) \z
@item \z
ASCII(26) (Control-Z).  This character can be encoded to allow you to
work around the problem that ASCII(26) stands for END-OF-FILE on Windows.
(ASCII(26) will cause problems if you try to use
@code{mysql database < filename}.)

@findex \\ (escape)
@findex escape (\\)
@item \\
A backslash (@samp{\}) character.

@findex % (wildcard character)
@findex Wildcard character (%)
@item \%
A @samp{%} character. This is used to search for literal instances of
@samp{%} in contexts where @samp{%} would otherwise be interpreted
as a wildcard character. @xref{String comparison functions}.

@findex _ (wildcard character)
@findex Wildcard character (_)
@item \_
A @samp{_} character. This is used to search for literal instances of
@samp{_} in contexts where @samp{_} would otherwise be interpreted
as a wildcard character. @xref{String comparison functions}.
@end table

Note that if you use @samp{\%} or @samp{\_} in some string contexts, these
will return the strings @samp{\%} and @samp{\_} and not @samp{%} and
@samp{_}.

@cindex quotes, in strings
@noindent
There are several ways to include quotes within a string:

@itemize @bullet
@item
A @samp{'} inside a string quoted with @samp{'} may be written as @samp{''}.

@item
A @samp{"} inside a string quoted with @samp{"} may be written as @samp{""}.

@item
You can precede the quote character with an escape character (@samp{\}).

@item
A @samp{'} inside a string quoted with @samp{"} needs no special treatment
and need not be doubled or escaped.  In the same way, @samp{"} inside a
string quoted with @samp{'} needs no special treatment.
@end itemize

The @code{SELECT} statements shown here demonstrate how quoting and
escaping work:

@example
mysql> SELECT 'hello', '"hello"', '""hello""', 'hel''lo', '\'hello';
+-------+---------+-----------+--------+--------+
| hello | "hello" | ""hello"" | hel'lo | 'hello |
+-------+---------+-----------+--------+--------+

mysql> SELECT "hello", "'hello'", "''hello''", "hel""lo", "\"hello";
+-------+---------+-----------+--------+--------+
| hello | 'hello' | ''hello'' | hel"lo | "hello |
+-------+---------+-----------+--------+--------+

mysql> SELECT "This\nIs\nFour\nlines";
+--------------------+
| This
Is
Four
lines |
+--------------------+
@end example

@cindex quoting binary data

If you want to insert binary data into a string column (such as a
@code{BLOB}), the following characters must be represented by escape
sequences:
@table @code
@item NUL
ASCII 0. You should represent this by @samp{\0} (a backslash and an ASCII @samp{0} character).
@item \
ASCII 92, backslash.  Represent this by @samp{\\}.
@item '
ASCII 39, single quote.  Represent this by @samp{\'}.
@item "
ASCII 34, double quote.  Represent this by @samp{\"}.
@end table

@cindex quoting
@cindex @code{BLOB}, inserting binary data
@findex mysql_real_escape_string()
@findex DBI->quote
If you write C code, you can use the C API function
@code{mysql_real_escape_string()} to escape characters for the @code{INSERT}
statement.  @xref{C API function overview}.  In Perl, you can use the
@code{quote} method of the @code{DBI} package to convert special
characters to the proper escape sequences.  @xref{Perl DBI Class, , Perl
@code{DBI} Class}.

You should use an escape function on any string that might contain any of the
special characters listed above!

Alternatively, many MySQL APIs provide some sort of placeholder capability
that allows you to insert special markers into a query string, and then bind
data values to them when you issue the query.  In this case, the API takes
case of escaping special characters in the values for you automatically.


@node Number syntax, Hexadecimal values, String syntax, Literals
@subsubsection Numbers

@cindex numbers
@cindex valid numbers, examples
@cindex integers
@cindex floats
@cindex negative values

Integers are represented as a sequence of digits. Floats use @samp{.} as a
decimal separator.  Either type of number may be preceded by @samp{-} to
indicate a negative value.

Examples of valid integers:

@example
1221
0
-32
@end example

Examples of valid floating-point numbers:

@example
294.42
-32032.6809e+10
148.00
@end example

An integer may be used in a floating-point context; it is interpreted
as the equivalent floating-point number.


@node Hexadecimal values, NULL values, Number syntax, Literals
@subsubsection Hexadecimal Values

@tindex hexadecimal values

MySQL supports hexadecimal values.  In numeric context these act
like an integer (64-bit precision). In string context these act like a binary
string where each pair of hex digits is converted to a character:

@example
mysql> SELECT x'FF'
         -> 255
mysql> SELECT 0xa+0;
         -> 10
mysql> SELECT 0x5061756c;
         -> Paul
@end example

The @code{x'hexstring'} syntax (new in 4.0) is based on ANSI SQL and the
@code{0x} syntax is based on ODBC.  Hexadecimal strings are often used by
ODBC to supply values for @code{BLOB} columns.
You can convert a string or a number to hexadecimal with the @code{HEX()}
function.

@node NULL values,  , Hexadecimal values, Literals
@subsubsection @code{NULL} Values

@tindex NULL value

The @code{NULL} value means ``no data'' and is different from values such
as @code{0} for numeric types or the empty string for string types.
@xref{Problems with NULL, , Problems with @code{NULL}}.

@code{NULL} may be represented by @code{\N} when using the text file import
or export formats (@code{LOAD DATA INFILE}, @code{SELECT ... INTO OUTFILE}).
@xref{LOAD DATA, , @code{LOAD DATA}}.


@node Legal names, Name case sensitivity, Literals, Language Structure
@subsection Database, Table, Index, Column, and Alias Names

@cindex names
@cindex legal names
@cindex databases, names
@cindex tables, names
@cindex indexes, names
@cindex columns, names
@cindex aliases, names

Database, table, index, column, and alias names all follow the same rules in
MySQL.

@tindex identifiers, quoting
@tindex quoting of identifiers
@tindex `
@tindex "
Note that the rules changed starting with MySQL Version 3.23.6 when we
introduced quoting of identifiers (database, table, and column names)
with @samp{`}. @samp{"} will also work to quote identifiers if you run
in ANSI mode. @xref{ANSI mode}.

@multitable @columnfractions .15 .15 .70
@item @strong{Identifier} @tab @strong{Max length} @tab @strong{Allowed characters}
@item Database @tab 64 @tab Any character that is allowed in a directory name except @samp{/} or @samp{.}.
@item Table @tab 64 @tab Any character that is allowed in a file name, except @samp{/} or @samp{.}.
@item Column @tab 64 @tab All characters.
@item Alias @tab 255 @tab All characters.
@end multitable

Note that in addition to the above, you can't have ASCII(0) or ASCII(255) or
the quoting character in an identifier.

Note that if the identifier is a restricted word or contains special characters
you must always quote it with @code{`} when you use it:

@example
mysql> SELECT * FROM `select` WHERE `select`.id > 100;
@end example

In previous versions of MySQL, the name rules are as follows:

@itemize @bullet
@item
A name may consist of alphanumeric characters from the current character set
and also @samp{_} and @samp{$}. The default character set is ISO-8859-1
Latin1; this may be changed with the @code{--default-character-set} option
to @code{mysqld}.
@xref{Character sets}.

@item
A name may start with any character that is legal in a name.  In particular,
a name may start with a digit (this differs from many other database
systems!).  However, a name cannot consist @emph{only} of digits.

@item
You cannot use the @samp{.} character in names because it is used to extend the
format by which you can refer to columns (see immediately below).
@end itemize

It is recommended that you do not use names like @code{1e}, because
an expression like @code{1e+1} is ambiguous. It may be interpreted as the
expression @code{1e + 1} or as the number @code{1e+1}.

In MySQL you can refer to a column using any of the following forms:

@multitable @columnfractions .30 .70
@item @strong{Column reference} @tab @strong{Meaning}
@item @code{col_name} @tab Column @code{col_name}
from whichever table used in the query contains a column of that name.
@item @code{tbl_name.col_name} @tab Column @code{col_name} from table
@code{tbl_name} of the current database.
@item @code{db_name.tbl_name.col_name} @tab Column @code{col_name} from table
@code{tbl_name} of the database @code{db_name}.  This form is available in
MySQL Version 3.22 or later.
@item
@code{`column_name`} @tab A column that is a keyword or contains special characters.
@end multitable

You need not specify a @code{tbl_name} or @code{db_name.tbl_name} prefix for
a column reference in a statement unless the reference would be ambiguous.
For example, suppose tables @code{t1} and @code{t2} each contain a column
@code{c}, and you retrieve @code{c} in a @code{SELECT} statement that uses
both @code{t1} and @code{t2}.  In this case, @code{c} is ambiguous because it
is not unique among the tables used in the statement, so you must indicate
which table you mean by writing @code{t1.c} or @code{t2.c}.  Similarly, if
you are retrieving from a table @code{t} in database @code{db1} and from a
table @code{t} in database @code{db2}, you must refer to columns in those
tables as @code{db1.t.col_name} and @code{db2.t.col_name}.

@cindex ODBC compatibility
@cindex compatibility, with ODBC
The syntax @code{.tbl_name} means the table @code{tbl_name} in the current
database.  This syntax is accepted for ODBC compatibility, because some ODBC
programs prefix table names with a @samp{.} character.


@node Name case sensitivity, Variables, Legal names, Language Structure
@subsection Case Sensitivity in Names

@cindex names, case-sensitivity
@cindex case-sensitivity, in names
@cindex database names, case sensitivity
@cindex table names, case sensitivity
@cindex column names, case sensitivity
@cindex alias names, case sensitivity

In MySQL, databases and tables correspond to directories and files
within those directories.  Consequently, the case sensitivity of the
underlying operating system determines the case sensitivity of database and
table names.  This means database and table names are case-insensitive in
Windows, and case-sensitive in most varieties of Unix (Mac OS X being an
exception).
@xref{Extensions to ANSI}.

@strong{Note}: although database and table names are case-insensitive for
Windows, you should not refer to a given database or table using different
cases within the same query.  The following query would not work because it
refers to a table both as @code{my_table} and as @code{MY_TABLE}:

@example
mysql> SELECT * FROM my_table WHERE MY_TABLE.col=1;
@end example

Column names and column aliases are case-insensitive in all cases.

Aliases on tables are case-sensitive.  The following query would not work
because it refers to the alias both as @code{a} and as @code{A}:

@example
mysql> SELECT col_name FROM tbl_name AS a
    ->                 WHERE a.col_name = 1 OR A.col_name = 2;
@end example

If you have trouble remembering the lettercase for database and table names,
adopt a consistent convention, such as always creating databases and
tables using lowercase names.

One way to avoid this problem is to start @code{mysqld} with @code{-O
lower_case_table_names=1}. By default this option is 1 on Windows and 0 on
Unix.

If @code{lower_case_table_names} is 1 MySQL will convert all
table names to lower case on storage and lookup.  Note that if you
change this option, you need to first convert your old table names to
lower case before starting @code{mysqld}.

If you move @code{MyISAM} files from a Windows to a *nix disk, you may
in some cases need to use the @file{mysql_fix_extensions} tool to fix-up
the case of the file extensions in each specified database directory
(lowercase @file{.frm}, uppercase @file{.MYI} and @file{.MYD}).
@file{mysql_fix_extensions} can be found in the @file{script} subdirectory.


@node Variables, Comments, Name case sensitivity, Language Structure
@subsection User Variables

@cindex variables, user
@cindex user variables
@cindex names, variables

MySQL supports thread-specific variables with the
@code{@@variablename} syntax.  A variable name may consist of
alphanumeric characters from the current character set and also
@samp{_}, @samp{$}, and @samp{.} . The default character set is
ISO-8859-1 Latin1; this may be changed with the
@code{--default-character-set} option to @code{mysqld}. @xref{Character
sets}.

Variables don't have to be initialised.  They contain @code{NULL} by default
and can store an integer, real, or string value.  All variables for
a thread are automatically freed when the thread exits.

You can set a variable with the @code{SET} syntax:

@example
SET @@variable= @{ integer expression | real expression | string expression @}
[,@@variable= ...].
@end example

You can also assign a value to a variable in statements other than @code{SET}.
However, in this case the assignment operator is @code{:=} rather than
@code{=}, because @code{=} is reserved for comparisons in non-@code{SET}
statements:

@example
mysql> SELECT @@t1:=(@@t2:=1)+@@t3:=4,@@t1,@@t2,@@t3;
+----------------------+------+------+------+
| @@t1:=(@@t2:=1)+@@t3:=4 | @@t1  | @@t2  | @@t3  |
+----------------------+------+------+------+
|                    5 |    5 |    1 |    4 |
+----------------------+------+------+------+
@end example

User variables may be used where expressions are allowed.  Note that
this does not currently include contexts where a number is explicitly
required, such as in the @code{LIMIT} clause of a @code{SELECT} statement,
or the @code{IGNORE number LINES} clause of a @code{LOAD DATA} statement.

@strong{Note}: in a @code{SELECT} statement, each expression is evaluated
only when it's sent to the client.  This means that in the @code{HAVING},
@code{GROUP BY}, or @code{ORDER BY} clause, you can't refer to an expression
that involves variables that are set in the @code{SELECT} part. For example,
the following statement will NOT work as expected:

@example
mysql> SELECT (@@aa:=id) AS a, (@@aa+3) AS b FROM table_name HAVING b=5;
@end example

The reason is that @code{@@aa} will not contain the value of the current
row, but the value of @code{id} for the previous accepted row.

@node Comments, Reserved words, Variables, Language Structure
@subsection Comment Syntax

@findex Comment syntax

@cindex comments, adding

The MySQL server supports the @code{# to end of line}, @code{--
to end of line} and @code{/* in-line or multiple-line */} comment
styles:

@example
mysql> SELECT 1+1;     # This comment continues to the end of line
mysql> SELECT 1+1;     -- This comment continues to the end of line
mysql> SELECT 1 /* this is an in-line comment */ + 1;
mysql> SELECT 1+
/*
this is a
multiple-line comment
*/
1;
@end example

Note that the @code{--} (double-dash) comment style requires you to have at
least one space after the second dash!

Although the server understands the comment syntax just described,
there are some limitations on the way that the @code{mysql} client
parses @code{/* ... */} comments:

@itemize @bullet
@item
Single-quote and double-quote characters are taken to indicate the beginning
of a quoted string, even within a comment.  If the quote is not matched by a
second quote within the comment, the parser doesn't realise the comment has
ended.  If you are running @code{mysql} interactively, you can tell that it
has gotten confused like this because the prompt changes from @code{mysql>}
to @code{'>} or @code{">}.

@item
A semicolon is taken to indicate the end of the current SQL statement
and anything following it to indicate the beginning of the next statement.
@end itemize

These limitations apply both when you run @code{mysql} interactively
and when you put commands in a file and tell @code{mysql} to read its
input from that file with @code{mysql < some-file}.

MySQL supports the @samp{--} ANSI SQL comment style only if the second dash
is followed by a space. @xref{ANSI diff comments}.


@node Reserved words,  , Comments, Language Structure
@subsection Is MySQL Picky About Reserved Words?

@cindex keywords
@cindex reserved words, exceptions

A common problem stems from trying to create a table with column names that
use the names of datatypes or functions built into MySQL, such as
@code{TIMESTAMP} or @code{GROUP}.  You're allowed to do it (for example,
@code{ABS} is an allowed column name), but whitespace is not allowed between
a function name and the immediately following @samp{(} when using functions
whose names are also column names.

The following words are explicitly reserved in MySQL. Most of
them are forbidden by ANSI SQL92 as column and/or table names
(for example, @code{GROUP}).
A few are reserved because MySQL needs them and is
(currently) using a @code{yacc} parser:

@c START_OF_RESERVED_WORDS

@c Reserved word list updated Fri May  3 08:41:00 2002 by arjen.
@c To regenerate, use Support/update-reserved-words.pl.

@multitable @columnfractions .33 .33 .34
@item @strong{Word}
 @tab @strong{Word}
 @tab @strong{Word}
@item @code{ADD}
 @tab @code{ALL}
 @tab @code{ALTER}
@item @code{ANALYZE}
 @tab @code{AND}
 @tab @code{AS}
@item @code{ASC}
 @tab @code{AUTO_INCREMENT}
 @tab @code{BDB}
@item @code{BERKELEYDB}
 @tab @code{BETWEEN}
 @tab @code{BIGINT}
@item @code{BINARY}
 @tab @code{BLOB}
 @tab @code{BOTH}
@item @code{BY}
 @tab @code{CASCADE}
 @tab @code{CASE}
@item @code{CHANGE}
 @tab @code{CHAR}
 @tab @code{CHARACTER}
@item @code{COLUMN}
 @tab @code{COLUMNS}
 @tab @code{CONSTRAINT}
@item @code{CREATE}
 @tab @code{CROSS}
 @tab @code{CURRENT_DATE}
@item @code{CURRENT_TIME}
 @tab @code{CURRENT_TIMESTAMP}
 @tab @code{DATABASE}
@item @code{DATABASES}
 @tab @code{DAY_HOUR}
 @tab @code{DAY_MINUTE}
@item @code{DAY_SECOND}
 @tab @code{DEC}
 @tab @code{DECIMAL}
@item @code{DEFAULT}
 @tab @code{DELAYED}
 @tab @code{DELETE}
@item @code{DESC}
 @tab @code{DESCRIBE}
 @tab @code{DISTINCT}
@item @code{DISTINCTROW}
 @tab @code{DOUBLE}
 @tab @code{DROP}
@item @code{ELSE}
 @tab @code{ENCLOSED}
 @tab @code{ESCAPED}
@item @code{EXISTS}
 @tab @code{EXPLAIN}
 @tab @code{FIELDS}
@item @code{FLOAT}
 @tab @code{FOR}
 @tab @code{FOREIGN}
@item @code{FROM}
 @tab @code{FULLTEXT}
 @tab @code{FUNCTION}
@item @code{GRANT}
 @tab @code{GROUP}
 @tab @code{HAVING}
@item @code{HIGH_PRIORITY}
 @tab @code{HOUR_MINUTE}
 @tab @code{HOUR_SECOND}
@item @code{IF}
 @tab @code{IGNORE}
 @tab @code{IN}
@item @code{INDEX}
 @tab @code{INFILE}
 @tab @code{INNER}
@item @code{INNODB}
 @tab @code{INSERT}
 @tab @code{INSERT_ID}
@item @code{INT}
 @tab @code{INTEGER}
 @tab @code{INTERVAL}
@item @code{INTO}
 @tab @code{IS}
 @tab @code{JOIN}
@item @code{KEY}
 @tab @code{KEYS}
 @tab @code{KILL}
@item @code{LAST_INSERT_ID}
 @tab @code{LEADING}
 @tab @code{LEFT}
@item @code{LIKE}
 @tab @code{LIMIT}
 @tab @code{LINES}
@item @code{LOAD}
 @tab @code{LOCK}
 @tab @code{LONG}
@item @code{LONGBLOB}
 @tab @code{LONGTEXT}
 @tab @code{LOW_PRIORITY}
@item @code{MASTER_SERVER_ID}
 @tab @code{MATCH}
 @tab @code{MEDIUMBLOB}
@item @code{MEDIUMINT}
 @tab @code{MEDIUMTEXT}
 @tab @code{MIDDLEINT}
@item @code{MINUTE_SECOND}
 @tab @code{MRG_MYISAM}
 @tab @code{NATURAL}
@item @code{NOT}
 @tab @code{NULL}
 @tab @code{NUMERIC}
@item @code{ON}
 @tab @code{OPTIMIZE}
 @tab @code{OPTION}
@item @code{OPTIONALLY}
 @tab @code{OR}
 @tab @code{ORDER}
@item @code{OUTER}
 @tab @code{OUTFILE}
 @tab @code{PARTIAL}
@item @code{PRECISION}
 @tab @code{PRIMARY}
 @tab @code{PRIVILEGES}
@item @code{PROCEDURE}
 @tab @code{PURGE}
 @tab @code{READ}
@item @code{REAL}
 @tab @code{REFERENCES}
 @tab @code{REGEXP}
@item @code{RENAME}
 @tab @code{REPLACE}
 @tab @code{REQUIRE}
@item @code{RESTRICT}
 @tab @code{RETURNS}
 @tab @code{REVOKE}
@item @code{RIGHT}
 @tab @code{RLIKE}
 @tab @code{SELECT}
@item @code{SET}
 @tab @code{SHOW}
 @tab @code{SMALLINT}
@item @code{SONAME}
 @tab @code{SQL_AUTO_IS_NULL}
 @tab @code{SQL_BIG_RESULT}
@item @code{SQL_BIG_SELECTS}
 @tab @code{SQL_BIG_TABLES}
 @tab @code{SQL_BUFFER_RESULT}
@item @code{SQL_CALC_FOUND_ROWS}
 @tab @code{SQL_LOG_BIN}
 @tab @code{SQL_LOG_OFF}
@item @code{SQL_LOG_UPDATE}
 @tab @code{SQL_LOW_PRIORITY_UPDATES}
 @tab @code{SQL_MAX_JOIN_SIZE}
@item @code{SQL_QUOTE_SHOW_CREATE}
 @tab @code{SQL_SAFE_UPDATES}
 @tab @code{SQL_SELECT_LIMIT}
@item @code{SQL_SLAVE_SKIP_COUNTER}
 @tab @code{SQL_SMALL_RESULT}
 @tab @code{SQL_WARNINGS}
@item @code{SSL}
 @tab @code{STARTING}
 @tab @code{STRAIGHT_JOIN}
@item @code{STRIPED}
 @tab @code{TABLE}
 @tab @code{TABLES}
@item @code{TERMINATED}
 @tab @code{THEN}
 @tab @code{TINYBLOB}
@item @code{TINYINT}
 @tab @code{TINYTEXT}
 @tab @code{TO}
@item @code{TRAILING}
 @tab @code{UNION}
 @tab @code{UNIQUE}
@item @code{UNLOCK}
 @tab @code{UNSIGNED}
 @tab @code{UPDATE}
@item @code{USAGE}
 @tab @code{USE}
 @tab @code{USING}
@item @code{VALUES}
 @tab @code{VARBINARY}
 @tab @code{VARCHAR}
@item @code{VARYING}
 @tab @code{WHEN}
 @tab @code{WHERE}
@item @code{WITH}
 @tab @code{WRITE}
 @tab @code{YEAR_MONTH}
@item @code{ZEROFILL}
 @tab
 @tab
@end multitable

@c END_OF_RESERVED_WORDS

The following symbols (from the table above) are disallowed by ANSI SQL
but allowed by MySQL as column/table names. This is because some
of these names are very natural names and a lot of people have already
used them.

@itemize @bullet
@item @code{ACTION}
@item @code{BIT}
@item @code{DATE}
@item @code{ENUM}
@item @code{NO}
@item @code{TEXT}
@item @code{TIME}
@item @code{TIMESTAMP}
@end itemize


@node Column types, Functions, Language Structure, Reference
@section Column Types

@cindex columns, types
@cindex types, columns

MySQL supports a number of column types, which may be grouped into
three categories: numeric types, date and time types, and string (character)
types.  This section first gives an overview of the types available and
summarises the storage requirements for each column type, then provides a
more detailed description of the properties of the types in each category.
The overview is intentionally brief.  The more detailed descriptions should
be consulted for additional information about particular column types, such
as the allowable formats in which you can specify values.

The column types supported by MySQL are listed below.
The following code letters are used in the descriptions:

@cindex display size
@cindex sizes, display
@cindex digits
@cindex decimal point
@cindex brackets, square
@cindex square brackets
@table @code
@item M
Indicates the maximum display size.  The maximum legal display size is 255.

@item D
Applies to floating-point types and indicates the number of digits
following the decimal point.  The maximum possible value is 30, but
should be no greater than @code{M}-2.
@end table

Square brackets (@samp{[} and @samp{]}) indicate parts of type specifiers
that are optional.

@tindex Types

@c The @w{-number} stuff keeps a linebreak from occurring between
@c the - and number.

Note that if you specify @code{ZEROFILL} for a column, MySQL will
automatically add the @code{UNSIGNED} attribute to the column.

@strong{Warning}: you should be aware that when you use subtraction
between integer values where one is of type @code{UNSIGNED}, the result
will be unsigned! @xref{Cast Functions}.

@table @code
@tindex TINYINT
@item TINYINT[(M)] [UNSIGNED] [ZEROFILL]

A very small integer. The signed range is @code{-128} to @code{127}. The
unsigned range is @code{0} to @code{255}.

@tindex BOOL
@tindex BIT
@item BIT
@itemx BOOL
These are synonyms for @code{TINYINT(1)}.

@tindex SMALLINT
@item SMALLINT[(M)] [UNSIGNED] [ZEROFILL]

A small integer. The signed range is @code{-32768} to @code{32767}. The
unsigned range is @code{0} to @code{65535}.

@tindex MEDIUMINT
@item MEDIUMINT[(M)] [UNSIGNED] [ZEROFILL]

A medium-size integer. The signed range is @code{-8388608} to
@code{8388607}. The unsigned range is @code{0} to @code{16777215}.

@tindex INT
@item INT[(M)] [UNSIGNED] [ZEROFILL]

A normal-size integer. The signed range is @code{-2147483648} to
@code{2147483647}.  The unsigned range is @code{0} to @code{4294967295}.

@tindex INTEGER
@item INTEGER[(M)] [UNSIGNED] [ZEROFILL]

This is a synonym for @code{INT}.

@tindex BIGINT
@item BIGINT[(M)] [UNSIGNED] [ZEROFILL]

A large integer. The signed range is @code{-9223372036854775808} to
@code{9223372036854775807}. The unsigned range is @code{0} to
@code{18446744073709551615}.

Some things you should be aware of with respect to @code{BIGINT} columns:

@itemize @bullet
@item
@cindex rounding errors
All arithmetic is done using signed @code{BIGINT} or @code{DOUBLE}
values, so you shouldn't use unsigned big integers larger than
@code{9223372036854775807} (63 bits) except with bit functions! If you
do that, some of the last digits in the result may be wrong because of
rounding errors when converting the @code{BIGINT} to a @code{DOUBLE}.

MySQL 4.0 can handle @code{BIGINT} in the following cases:
@itemize @bullet
@item
Use integers to store big unsigned values in a @code{BIGINT} column.
@item
In @code{MIN(big_int_column)} and @code{MAX(big_int_column)}.
@item
When using operators (@code{+}, @code{-}, @code{*}, etc.) where
both operands are integers.
@end itemize

@item
You can always store an exact integer value in a @code{BIGINT} column by
storing it as a string.  In this case, MySQL will perform a string-to-number
conversion that involves no intermediate double representation.
@item
@samp{-}, @samp{+}, and @samp{*} will use @code{BIGINT} arithmetic when
both arguments are integer values!  This means that if you
multiply two big integers (or results from functions that return
integers) you may get unexpected results when the result is larger than
@code{9223372036854775807}.
@end itemize

@cindex floating-point number
@tindex FLOAT
@tindex FLOAT(precision)
@item FLOAT(precision) [UNSIGNED] [ZEROFILL]

A floating-point number.  @code{precision} can be
@code{<=24} for a single-precision floating-point number and between 25
and 53 for a double-precision floating-point number. These types are like
the @code{FLOAT} and @code{DOUBLE} types described immediately below.
@code{FLOAT(X)} has the same range as the corresponding @code{FLOAT} and
@code{DOUBLE} types, but the display size and number of decimals are undefined.

In MySQL Version 3.23, this is a true floating-point value.  In
earlier MySQL versions, @code{FLOAT(precision)} always has 2 decimals.

Note that using @code{FLOAT} may give you some unexpected problems as
all calculations in MySQL are done with double precision.
@xref{No matching rows}.

@cindex ODBC compatibility
@cindex compatibility, with ODBC
This syntax is provided for ODBC compatibility.

@tindex FLOAT
@tindex FLOAT(M,D)
@item FLOAT[(M,D)] [UNSIGNED] [ZEROFILL]

A small (single-precision) floating-point number.  Allowable values are
@code{@w{-3.402823466E+38}} to @code{@w{-1.175494351E-38}}, @code{0},
and @code{@w{1.175494351E-38}} to @code{3.402823466E+38}.  If
@code{UNSIGNED} is specified, negative values are disallowed.  The @code{M}
is the display width and @code{D} is the number of decimals.  @code{FLOAT}
without arguments or @code{FLOAT(X)} where @code{X} <= 24 stands for a
single-precision floating-point number.

@tindex DOUBLE
@tindex FLOAT(precision)
@item DOUBLE[(M,D)] [UNSIGNED] [ZEROFILL]

A normal-size (double-precision) floating-point number.
Allowable values are @code{@w{-1.7976931348623157E+308}} to
@code{@w{-2.2250738585072014E-308}}, @code{0}, and
@code{2.2250738585072014E-308} to @code{1.7976931348623157E+308}.  If
@code{UNSIGNED} is specified, negative values are disallowed.  The
@code{M} is the display width and @code{D} is the number of decimals.
@code{DOUBLE} without arguments or @code{FLOAT(X)} where 25 <= @code{X}
<= 53 stands for a double-precision floating-point number.

@tindex DOUBLE PRECISION
@tindex REAL
@item DOUBLE PRECISION[(M,D)] [UNSIGNED] [ZEROFILL]
@itemx REAL[(M,D)] [UNSIGNED] [ZEROFILL]

These are synonyms for @code{DOUBLE}.

@tindex DECIMAL
@item DECIMAL[(M[,D])] [UNSIGNED] [ZEROFILL]

An unpacked floating-point number.  Behaves like a
@code{CHAR} column: ``unpacked'' means the number is stored as a string,
using one character for each digit of the value.  The decimal point and,
for negative numbers, the @samp{-} sign, are not counted in @code{M} (but
space for these is reserved). If @code{D} is 0, values will have no decimal
point or fractional part.  The maximum range of @code{DECIMAL} values is
the same as for @code{DOUBLE}, but the actual range for a given
@code{DECIMAL} column may be constrained by the choice of @code{M} and
@code{D}.  If @code{UNSIGNED} is specified, negative values are disallowed.  

If @code{D} is omitted, the default is 0.  If @code{M} is omitted, the
default is 10.

Prior to MySQL Version 3.23, the @code{M} argument must include the space
needed for the sign and the decimal point.

@tindex DEC
@tindex NUMERIC
@item DEC[(M[,D])] [UNSIGNED] [ZEROFILL]
@itemx NUMERIC[(M[,D])] [UNSIGNED] [ZEROFILL]

These are synonyms for @code{DECIMAL}.

@tindex DATE
@item DATE

A date.  The supported range is @code{'1000-01-01'} to @code{'9999-12-31'}.
MySQL displays @code{DATE} values in @code{'YYYY-MM-DD'} format, but
allows you to assign values to @code{DATE} columns using either strings or
numbers. @xref{DATETIME}.

@tindex DATETIME
@item DATETIME

A date and time combination.  The supported range is @code{'1000-01-01
00:00:00'} to @code{'9999-12-31 23:59:59'}.  MySQL displays
@code{DATETIME} values in @code{'YYYY-MM-DD HH:MM:SS'} format, but allows you
to assign values to @code{DATETIME} columns using either strings or numbers.
@xref{DATETIME}.

@tindex TIMESTAMP
@item TIMESTAMP[(M)]

A timestamp.  The range is @code{'1970-01-01 00:00:00'} to sometime in the
year @code{2037}.  MySQL displays @code{TIMESTAMP} values in
@code{YYYYMMDDHHMMSS}, @code{YYMMDDHHMMSS}, @code{YYYYMMDD}, or @code{YYMMDD}
format, depending on whether @code{M} is @code{14} (or missing), @code{12},
@code{8}, or @code{6}, but allows you to assign values to @code{TIMESTAMP}
columns using either strings or numbers.  A @code{TIMESTAMP} column is useful
for recording the date and time of an @code{INSERT} or @code{UPDATE}
operation because it is automatically set to the date and time of the most
recent operation if you don't give it a value yourself.  You can also set it
to the current date and time by assigning it a @code{NULL} value.  @xref{Date
and time types}.

The @code{M} argument affects only how a @code{TIMESTAMP} column is displayed;
its values always are stored using 4 bytes each.

Note that @code{TIMESTAMP(M)} columns where @code{M} is 8 or 14 are reported to
be numbers while other @code{TIMESTAMP(M)} columns are reported to be
strings.  This is just to ensure that one can reliably dump and restore
the table with these types!
@xref{DATETIME, , @code{DATETIME}}.

@tindex TIME
@item TIME

A time.  The range is @code{'-838:59:59'} to @code{'838:59:59'}.
MySQL displays @code{TIME} values in @code{'HH:MM:SS'} format, but
allows you to assign values to @code{TIME} columns using either strings or
numbers. @xref{TIME, , @code{TIME}}.

@tindex YEAR
@item YEAR[(2|4)]

A year in 2- or 4-digit format (default is 4-digit).  The allowable values
are @code{1901} to @code{2155}, @code{0000} in the 4-digit year format,
and 1970-2069 if you use the 2-digit format (70-69).  MySQL displays
@code{YEAR} values in @code{YYYY} format, but allows you to assign values to
@code{YEAR} columns using either strings or numbers. (The @code{YEAR} type is
unavailable prior to MySQL Version 3.22.) @xref{YEAR, , @code{YEAR}}.

@tindex NATIONAL CHAR
@tindex NCHAR
@tindex CHAR
@tindex CHARACTER
@item [NATIONAL] CHAR(M) [BINARY]

A fixed-length string that is always right-padded with spaces to the
specified length when stored. The range of @code{M} is 0 to 255 characters
(1 to 255 prior to MySQL Version 3.23).
Trailing spaces are removed when the value is retrieved. @code{CHAR} values
are sorted and compared in case-insensitive fashion according to the
default character set unless the @code{BINARY} keyword is given.

@code{NATIONAL CHAR} (or its equivalent short form, @code{NCHAR}) is the
ANSI SQL way to define that a @code{CHAR} column should use the default
CHARACTER set.  This is the default in MySQL.

@code{CHAR} is a shorthand for @code{CHARACTER}.

MySQL allows you to create a column of type
@code{CHAR(0)}. This is mainly useful when you have to be compliant with
some old applications that depend on the existence of a column but that do not
actually use the value.  This is also quite nice when you need a
column that only can take 2 values: A @code{CHAR(0)}, that is not defined
as @code{NOT NULL}, will occupy only one bit and can take only 2 values:
@code{NULL} or @code{""}. @xref{CHAR, , @code{CHAR}}.

@item CHAR
This is a synonym for @code{CHAR(1)}.

@tindex CHARACTER VARYING
@tindex CHAR VARYING
@tindex VARCHAR
@item [NATIONAL] VARCHAR(M) [BINARY]

A variable-length string.  @strong{Note}: trailing spaces are removed when
the value is stored (this differs from the ANSI SQL specification). The range
of @code{M} is 0 to 255 characters (1 to 255 prior to MySQL Version 4.0.2).
@code{VARCHAR} values are sorted and compared in case-insensitive fashion
unless the @code{BINARY} keyword is given. @xref{Silent column changes}.

@code{VARCHAR} is a shorthand for @code{CHARACTER VARYING}.
@xref{CHAR}.

@tindex TINYBLOB
@tindex TINYTEXT
@item TINYBLOB
@itemx TINYTEXT

A @code{BLOB} or @code{TEXT} column with a maximum length of 255 (2^8 - 1)
characters. @xref{Silent column changes}. @xref{BLOB}.

@tindex BLOB
@tindex TEXT
@item BLOB
@itemx TEXT

A @code{BLOB} or @code{TEXT} column with a maximum length of 65535 (2^16 - 1)
characters. @xref{Silent column changes}. @xref{BLOB}.

@tindex MEDIUMBLOB
@tindex MEDIUMTEXT
@item MEDIUMBLOB
@itemx MEDIUMTEXT

A @code{BLOB} or @code{TEXT} column with a maximum length of 16777215
(2^24 - 1) characters. @xref{Silent column changes}. @xref{BLOB}.

@tindex LONGBLOB
@tindex LONGTEXT
@item LONGBLOB
@itemx LONGTEXT

A @code{BLOB} or @code{TEXT} column with a maximum length of 4294967295
(2^32 - 1) characters. @xref{Silent column changes}.  Note that because
the server/client protocol and MyISAM tables has currently a limit of
16M per communication packet / table row, you can't yet use this
the whole range of this type. @xref{BLOB}.

@tindex ENUM
@item ENUM('value1','value2',...)

An enumeration.  A string object that can have only one value, chosen
from the list of values @code{'value1'}, @code{'value2'}, @code{...},
@code{NULL} or the special @code{""} error value.  An @code{ENUM} can
have a maximum of 65535 distinct values. @xref{ENUM}.

@tindex SET
@item SET('value1','value2',...)

A set.  A string object that can have zero or more values, each of which must
be chosen from the list of values @code{'value1'}, @code{'value2'},
@code{...} A @code{SET} can have a maximum of 64 members. @xref{SET}.
@end table

@menu
* Numeric types::               Numeric types
* Date and time types::         Date and time types
* String types::                String types
* Choosing types::              Choosing the right type for a column
* Other-vendor column types::   Using column types from other database engines
* Storage requirements::        Column type storage requirements
@end menu


@node Numeric types, Date and time types, Column types, Column types
@subsection Numeric Types

MySQL supports all of the ANSI/ISO SQL92 numeric types.  These
types include the exact numeric data types (@code{NUMERIC},
@code{DECIMAL}, @code{INTEGER}, and @code{SMALLINT}), as well as the
approximate numeric data types (@code{FLOAT}, @code{REAL}, and
@code{DOUBLE PRECISION}). The keyword @code{INT} is a synonym for
@code{INTEGER}, and the keyword @code{DEC} is a synonym for
@code{DECIMAL}.

The @code{NUMERIC} and @code{DECIMAL} types are implemented as the same
type by MySQL, as permitted by the SQL92 standard.  They are
used for values for which it is important to preserve exact precision,
for example with monetary data.  When declaring a column of one of these
types the precision and scale can be (and usually is) specified; for
example:

@example
    salary DECIMAL(9,2)
@end example

In this example, @code{9} (@code{precision}) represents the number of
significant decimal digits that will be stored for values, and @code{2}
(@code{scale}) represents the number of digits that will be stored
following the decimal point.  In this case, therefore, the range of
values that can be stored in the @code{salary} column is from
@code{-9999999.99} to @code{9999999.99}.
(MySQL can actually store numbers up to @code{9999999.99} in this column
because it doesn't have to store the sign for positive numbers)

In ANSI/ISO SQL92, the syntax @code{DECIMAL(p)} is equivalent to
@code{DECIMAL(p,0)}.  Similarly, the syntax @code{DECIMAL} is equivalent
to @code{DECIMAL(p,0)}, where the implementation is allowed to decide
the value of @code{p}.  MySQL does not currently support either of these
variant forms of the @code{DECIMAL}/@code{NUMERIC} data types.  This is
not generally a serious problem, as the principal benefits of these
types derive from the ability to control both precision and scale
explicitly.

@code{DECIMAL} and @code{NUMERIC} values are stored as strings, rather
than as binary floating-point numbers, in order to preserve the decimal
precision of those values.  One character is used for each digit of the
value, the decimal point (if @code{scale} > 0), and the @samp{-} sign
(for negative numbers).  If @code{scale} is 0, @code{DECIMAL} and
@code{NUMERIC} values contain no decimal point or fractional part.

The maximum range of @code{DECIMAL} and @code{NUMERIC} values is the
same as for @code{DOUBLE}, but the actual range for a given
@code{DECIMAL} or @code{NUMERIC} column can be constrained by the
@code{precision} or @code{scale} for a given column.  When such a column
is assigned a value with more digits following the decimal point than
are allowed by the specified @code{scale}, the value is rounded to that
@code{scale}.  When a @code{DECIMAL} or @code{NUMERIC} column is
assigned a value whose magnitude exceeds the range implied by the
specified (or defaulted) @code{precision} and @code{scale},
MySQL stores the value representing the corresponding end
point of that range.

As an extension to the ANSI/ISO SQL92 standard, MySQL also
supports the integer types @code{TINYINT}, @code{MEDIUMINT}, and
@code{BIGINT} as listed in the tables above.  Another extension is
supported by MySQL for optionally specifying the display width
of an integer value in parentheses following the base keyword for the
type (for example, @code{INT(4)}).  This optional width specification is
used to left-pad the display of values whose width is less than the
width specified for the column, but does not constrain the range of
values that can be stored in the column, nor the number of digits that
will be displayed for values whose width exceeds that specified for the
column.  When used in conjunction with the optional extension attribute
@code{ZEROFILL}, the default padding of spaces is replaced with zeroes.
For example, for a column declared as @code{INT(5) ZEROFILL}, a value
of @code{4} is retrieved as @code{00004}.  Note that if you store larger
values than the display width in an integer column, you may experience
problems when MySQL generates temporary tables for some
complicated joins, as in these cases MySQL trusts that the
data did fit into the original column width.

All integer types can have an optional (non-standard) attribute
@code{UNSIGNED}.  Unsigned values can be used when you want to allow
only positive numbers in a column and you need a little bigger numeric
range for the column.

As of MySQL 4.0.2, floating-point types also can be @code{UNSIGNED}.
As with integer types, this attribute prevents negative values from
being stored in the column.  Unlike the integer types, the upper range
of column values remains the same.

The @code{FLOAT} type is used to represent approximate numeric data
types. The ANSI/ISO SQL92 standard allows an optional specification of
the precision (but not the range of the exponent) in bits following the
keyword @code{FLOAT} in parentheses.  The MySQL implementation
also supports this optional precision specification.  When the keyword
@code{FLOAT} is used for a column type without a precision
specification, MySQL uses four bytes to store the values.  A
variant syntax is also supported, with two numbers given in parentheses
following the @code{FLOAT} keyword.  With this option, the first number
continues to represent the storage requirements for the value in bytes,
and the second number specifies the number of digits to be stored and
displayed following the decimal point (as with @code{DECIMAL} and
@code{NUMERIC}).  When MySQL is asked to store a number for
such a column with more decimal digits following the decimal point than
specified for the column, the value is rounded to eliminate the extra
digits when the value is stored.

The @code{REAL} and @code{DOUBLE PRECISION} types do not accept
precision specifications.  As an extension to the ANSI/ISO SQL92
standard, MySQL recognises @code{DOUBLE} as a synonym for the
@code{DOUBLE PRECISION} type.  In contrast with the standard's
requirement that the precision for @code{REAL} be smaller than that used
for @code{DOUBLE PRECISION}, MySQL implements both as 8-byte
double-precision floating-point values (when not running in ``ANSI mode'').
For maximum portability, code requiring storage of approximate numeric
data values should use @code{FLOAT} or @code{DOUBLE PRECISION} with no
specification of precision or number of decimal points.

When asked to store a value in a numeric column that is outside the column
type's allowable range, MySQL clips the value to the appropriate
endpoint of the range and stores the resulting value instead.

For example, the range of an @code{INT} column is @code{-2147483648} to
@code{2147483647}.  If you try to insert @code{-9999999999} into an
@code{INT} column, the value is clipped to the lower endpoint of the range,
and @code{-2147483648} is stored instead. Similarly, if you try to insert
@code{9999999999}, @code{2147483647} is stored instead.

If the @code{INT} column is @code{UNSIGNED}, the size of the column's
range is the same but its endpoints shift up to @code{0} and @code{4294967295}.
If you try to store @code{-9999999999} and @code{9999999999},
the values stored in the column become @code{0} and @code{4294967296}.

Conversions that occur due to clipping are reported as ``warnings'' for
@code{ALTER TABLE}, @code{LOAD DATA INFILE}, @code{UPDATE}, and
multi-row @code{INSERT} statements.


@node Date and time types, String types, Numeric types, Column types
@subsection Date and Time Types

@cindex types, Date and Time
@cindex Date and Time types

@menu
* Y2K issues::                  Y2K issues and date types
* DATETIME::                    The @code{DATETIME}, @code{DATE} and @code{TIMESTAMP} types
* TIME::                        The @code{TIME} type
* YEAR::                        The @code{YEAR} type
@end menu


The date and time types are @code{DATETIME}, @code{DATE},
@code{TIMESTAMP}, @code{TIME}, and @code{YEAR}.  Each of these has a
range of legal values, as well as a ``zero'' value that is used when you
specify a really illegal value.  Note that MySQL allows you to store
certain 'not strictly' legal date values, for example @code{1999-11-31}.
The reason for this is that we think it's the responsibility of the
application to handle date checking, not the SQL servers.  To make the
date checking 'fast', MySQL only checks that the month is in
the range of 0-12 and the day is in the range of 0-31. The above ranges
are defined this way because MySQL allows you to store, in a
@code{DATE} or @code{DATETIME} column, dates where the day or month-day
is zero.  This is extremely useful for applications that need to store
a birth-date for which you don't know the exact date. In this case you
simply store the date like @code{1999-00-00} or @code{1999-01-00}.  (You
cannot expect to get a correct value from functions like @code{DATE_SUB()}
or @code{DATE_ADD} for dates like these.)

Here are some general considerations to keep in mind when working
with date and time types:

@itemize @bullet
@item
MySQL retrieves values for a given date or time type in a standard
format, but it attempts to interpret a variety of formats for values that
you supply (for example, when you specify a value to be assigned to or
compared to a date or time type).  Nevertheless, only the formats described
in the following sections are supported.  It is expected that you will supply
legal values, and unpredictable results may occur if you use values in other
formats.

@item
Although MySQL tries to interpret values in several formats, it
always expects the year part of date values to be leftmost.  Dates must be
given in year-month-day order (for example, @code{'98-09-04'}), rather than
in the month-day-year or day-month-year orders commonly used elsewhere (for
example, @code{'09-04-98'}, @code{'04-09-98'}).

@item
MySQL automatically converts a date or time type value to a number
if the value is used in a numeric context, and vice versa.

@item
When MySQL encounters a value for a date or time type that is
out of range or otherwise illegal for the type (see the start of this
section), it converts the value to the ``zero'' value for that type.
(The exception is that out-of-range @code{TIME} values are clipped to
the appropriate endpoint of the @code{TIME} range.)  The following table
shows the format of the ``zero'' value for each type:

@multitable @columnfractions .15 .55
@item @strong{Column type} @tab @strong{``Zero'' value}
@item @code{DATETIME}      @tab @code{'0000-00-00 00:00:00'}
@item @code{DATE}          @tab @code{'0000-00-00'}
@item @code{TIMESTAMP}     @tab @code{00000000000000} (length depends on display size)
@item @code{TIME}          @tab @code{'00:00:00'}
@item @code{YEAR}          @tab @code{0000}
@end multitable

@item
The ``zero'' values are special, but you can store or refer to them
explicitly using the values shown in the table.  You can also do this
using the values @code{'0'} or @code{0}, which are easier to write.

@item
``Zero'' date or time values used through @code{MyODBC} are converted
automatically to @code{NULL} in @code{MyODBC} Version 2.50.12 and above,
because ODBC can't handle such values.
@end itemize


@node Y2K issues, DATETIME, Date and time types, Date and time types
@subsubsection Y2K Issues and Date Types

@cindex Year 2000 issues
@cindex date types, Y2K issues

MySQL itself is Y2K-safe (@pxref{Year 2000 compliance}),
but input values presented to MySQL may not be.  Any input
containing 2-digit year values is ambiguous, because the century is unknown.
Such values must be interpreted into 4-digit form because MySQL stores
years internally using four digits.

For @code{DATETIME}, @code{DATE}, @code{TIMESTAMP}, and @code{YEAR} types,
MySQL interprets dates with ambiguous year values using the
following rules:

@itemize @bullet
@item
Year values in the range @code{00-69} are converted to @code{2000-2069}.

@item
Year values in the range @code{70-99} are converted to @code{1970-1999}.
@end itemize

Remember that these rules provide only reasonable guesses as to what your
data mean.  If the heuristics used by MySQL don't produce the
correct values, you should provide unambiguous input containing 4-digit
year values.

@code{ORDER BY} will sort 2-digit @code{YEAR/DATE/DATETIME} types properly.

Note also that some functions like @code{MIN()} and @code{MAX()} will convert a
@code{TIMESTAMP/DATE} to a number. This means that a timestamp with a
2-digit year will not work properly with these functions.  The fix in this
case is to convert the @code{TIMESTAMP/DATE} to 4-digit year format or
use something like @code{MIN(DATE_ADD(timestamp,INTERVAL 0 DAYS))}.


@node DATETIME, TIME, Y2K issues, Date and time types
@subsubsection The @code{DATETIME}, @code{DATE}, and @code{TIMESTAMP} Types

@tindex DATETIME
@tindex DATE
@tindex TIMESTAMP

The @code{DATETIME}, @code{DATE}, and @code{TIMESTAMP} types are related.
This section describes their characteristics, how they are similar, and how
they differ.

The @code{DATETIME} type is used when you need values that contain both date
and time information.  MySQL retrieves and displays @code{DATETIME}
values in @code{'YYYY-MM-DD HH:MM:SS'} format.  The supported range is
@code{'1000-01-01 00:00:00'} to @code{'9999-12-31 23:59:59'}.  (``Supported''
means that although earlier values might work, there is no guarantee that
they will.)

The @code{DATE} type is used when you need only a date value, without a time
part.  MySQL retrieves and displays @code{DATE} values in
@code{'YYYY-MM-DD'} format.  The supported range is @code{'1000-01-01'} to
@code{'9999-12-31'}.

The @code{TIMESTAMP} column type provides a type that you can use to
automatically mark @code{INSERT} or @code{UPDATE} operations with the current
date and time.  If you have multiple @code{TIMESTAMP} columns, only the first
one is updated automatically.

Automatic updating of the first @code{TIMESTAMP} column occurs under any of
the following conditions:

@itemize @bullet
@item
The column is not specified explicitly in an @code{INSERT} or
@code{LOAD DATA INFILE} statement.
@item
The column is not specified explicitly in an @code{UPDATE} statement and some
other column changes value.  (Note that an @code{UPDATE} that sets a column
to the value it already has will not cause the @code{TIMESTAMP} column to be
updated, because if you set a column to its current value, MySQL
ignores the update for efficiency.)
@item
You explicitly set the @code{TIMESTAMP} column to @code{NULL}.
@end itemize

@code{TIMESTAMP} columns other than the first may also be set to the current
date and time.  Just set the column to @code{NULL} or to @code{NOW()}.

You can set any @code{TIMESTAMP} column to a value different from the current
date and time by setting it explicitly to the desired value.  This is true
even for the first @code{TIMESTAMP} column.  You can use this property if,
for example, you want a @code{TIMESTAMP} to be set to the current date and
time when you create a row, but not to be changed whenever the row is updated
later:

@itemize @bullet
@item
Let MySQL set the column when the row is created.
This will initialise it to the current date and time.

@item
When you perform subsequent updates to other columns in the row, set
the @code{TIMESTAMP} column explicitly to its current value.
@end itemize

On the other hand, you may find it just as easy to use a @code{DATETIME}
column that you initialise to @code{NOW()} when the row is created and
leave alone for subsequent updates.

@code{TIMESTAMP} values may range from the beginning of 1970 to sometime in
the year 2037, with a resolution of one second.  Values are displayed as
numbers.

The format in which MySQL retrieves and displays @code{TIMESTAMP}
values depends on the display size, as illustrated by the following table.  The
`full' @code{TIMESTAMP} format is 14 digits, but @code{TIMESTAMP} columns may
be created with shorter display sizes:

@multitable @columnfractions .20 .25
@item @strong{Column type}  @tab @strong{Display format}
@item @code{TIMESTAMP(14)}  @tab @code{YYYYMMDDHHMMSS}
@item @code{TIMESTAMP(12)}  @tab @code{YYMMDDHHMMSS}
@item @code{TIMESTAMP(10)}  @tab @code{YYMMDDHHMM}
@item @code{TIMESTAMP(8)}   @tab @code{YYYYMMDD}
@item @code{TIMESTAMP(6)}   @tab @code{YYMMDD}
@item @code{TIMESTAMP(4)}   @tab @code{YYMM}
@item @code{TIMESTAMP(2)}   @tab @code{YY}
@end multitable

All @code{TIMESTAMP} columns have the same storage size, regardless of
display size.  The most common display sizes are 6, 8, 12, and 14.  You can
specify an arbitrary display size at table creation time, but values of 0 or
greater than 14 are coerced to 14.  Odd-valued sizes in the range from 1 to
13 are coerced to the next higher even number.

You can specify @code{DATETIME}, @code{DATE}, and @code{TIMESTAMP} values using
any of a common set of formats:

@itemize @bullet
@item
As a string in either @code{'YYYY-MM-DD HH:MM:SS'} or @code{'YY-MM-DD
HH:MM:SS'} format.  A ``relaxed'' syntax is allowed---any punctuation
character may be used as the delimiter between date parts or time parts.
For example, @code{'98-12-31 11:30:45'}, @code{'98.12.31 11+30+45'},
@code{'98/12/31 11*30*45'}, and @code{'98@@12@@31 11^30^45'} are
equivalent.

@item
As a string in either @code{'YYYY-MM-DD'} or @code{'YY-MM-DD'} format.
A ``relaxed'' syntax is allowed here, too.  For example, @code{'98-12-31'},
@code{'98.12.31'}, @code{'98/12/31'}, and @code{'98@@12@@31'} are
equivalent.

@item
As a string with no delimiters in either @code{'YYYYMMDDHHMMSS'} or
@code{'YYMMDDHHMMSS'} format, provided that the string makes sense as a
date.  For example, @code{'19970523091528'} and @code{'970523091528'} are
interpreted as @code{'1997-05-23 09:15:28'}, but @code{'971122129015'} is
illegal (it has a nonsensical minute part) and becomes @code{'0000-00-00
00:00:00'}.

@item
As a string with no delimiters in either @code{'YYYYMMDD'} or @code{'YYMMDD'}
format, provided that the string makes sense as a date.  For example,
@code{'19970523'} and @code{'970523'} are interpreted as
@code{'1997-05-23'}, but @code{'971332'} is illegal (it has nonsensical month
and day parts) and becomes @code{'0000-00-00'}.

@item
As a number in either @code{YYYYMMDDHHMMSS} or @code{YYMMDDHHMMSS}
format, provided that the number makes sense as a date.  For example,
@code{19830905132800} and @code{830905132800} are interpreted as
@code{'1983-09-05 13:28:00'}.

@item
As a number in either @code{YYYYMMDD} or @code{YYMMDD}
format, provided that the number makes sense as a date.  For example,
@code{19830905} and @code{830905} are interpreted as @code{'1983-09-05'}.

@item
As the result of a function that returns a value that is acceptable
in a @code{DATETIME}, @code{DATE}, or @code{TIMESTAMP} context, such as
@code{NOW()} or @code{CURRENT_DATE}.
@end itemize

Illegal @code{DATETIME}, @code{DATE}, or @code{TIMESTAMP} values are converted
to the ``zero'' value of the appropriate type (@code{'0000-00-00 00:00:00'},
@code{'0000-00-00'}, or @code{00000000000000}).

For values specified as strings that include date part delimiters, it is not
necessary to specify two digits for month or day values that are less than
@code{10}.  @code{'1979-6-9'} is the same as @code{'1979-06-09'}.  Similarly,
for values specified as strings that include time part delimiters, it is not
necessary to specify two digits for hour, month, or second values that are
less than @code{10}.  @code{'1979-10-30 1:2:3'} is the same as
@code{'1979-10-30 01:02:03'}.

Values specified as numbers should be 6, 8, 12, or 14 digits long.  If the
number is 8 or 14 digits long, it is assumed to be in @code{YYYYMMDD} or
@code{YYYYMMDDHHMMSS} format and that the year is given by the first 4
digits.  If the number is 6 or 12 digits long, it is assumed to be in
@code{YYMMDD} or @code{YYMMDDHHMMSS} format and that the year is given by the
first 2 digits.  Numbers that are not one of these lengths are interpreted
as though padded with leading zeros to the closest length.

@cindex non-delimited strings
@cindex strings, non-delimited
Values specified as non-delimited strings are interpreted using their length
as given.  If the string is 8 or 14 characters long, the year is assumed to
be given by the first 4 characters.  Otherwise, the year is assumed to be
given by the first 2 characters.  The string is interpreted from left to
right to find year, month, day, hour, minute, and second values, for as many
parts as are present in the string.  This means you should not use strings
that have fewer than 6 characters.  For example, if you specify @code{'9903'},
thinking that will represent March, 1999, you will find that MySQL
inserts a ``zero'' date into your table.  This is because the year and month
values are @code{99} and @code{03}, but the day part is missing (zero), so
the value is not a legal date.

@code{TIMESTAMP} columns store legal values using the full precision with
which the value was specified, regardless of the display size.  This has
several implications:

@itemize @bullet
@item
Always specify year, month, and day, even if your column types are
@code{TIMESTAMP(4)} or @code{TIMESTAMP(2)}.  Otherwise, the value will not
be a legal date and @code{0} will be stored.

@item
If you use @code{ALTER TABLE} to widen a narrow @code{TIMESTAMP} column,
information will be displayed that previously was ``hidden''.

@item
Similarly, narrowing a @code{TIMESTAMP} column does not cause information to
be lost, except in the sense that less information is shown when the values
are displayed.

@item
Although @code{TIMESTAMP} values are stored to full precision, the only
function that operates directly on the underlying stored value is
@code{UNIX_TIMESTAMP()}.  Other functions operate on the formatted retrieved
value.  This means you cannot use functions such as @code{HOUR()} or
@code{SECOND()} unless the relevant part of the @code{TIMESTAMP} value is
included in the formatted value.  For example, the @code{HH} part of a
@code{TIMESTAMP} column is not displayed unless the display size is at least
10, so trying to use @code{HOUR()} on shorter @code{TIMESTAMP} values
produces a meaningless result.
@end itemize

You can to some extent assign values of one date type to an object
of a different date type.  However, there may be some alteration
of the value or loss of information:

@itemize @bullet
@item
If you assign a @code{DATE} value to a @code{DATETIME} or @code{TIMESTAMP}
object, the time part of the resulting value is set to @code{'00:00:00'},
because the @code{DATE} value contains no time information.

@item
If you assign a @code{DATETIME} or @code{TIMESTAMP} value to a @code{DATE}
object, the time part of the resulting value is deleted, because the
@code{DATE} type stores no time information.

@item
Remember that although @code{DATETIME}, @code{DATE}, and @code{TIMESTAMP}
values all can be specified using the same set of formats, the types do not
all have the same range of values.  For example, @code{TIMESTAMP} values
cannot be earlier than @code{1970} or later than @code{2037}.  This means
that a date such as @code{'1968-01-01'}, while legal as a @code{DATETIME} or
@code{DATE} value, is not a valid @code{TIMESTAMP} value and will be
converted to @code{0} if assigned to such an object.
@end itemize

@cindex problems, date values
@cindex date values, problems
Be aware of certain pitfalls when specifying date values:

@itemize @bullet
@item
The relaxed format allowed for values specified as strings can be deceiving.
For example, a value such as @code{'10:11:12'} might look like a time value
because of the @samp{:} delimiter, but if used in a date context will be
interpreted as the year @code{'2010-11-12'}.  The value @code{'10:45:15'}
will be converted to @code{'0000-00-00'} because @code{'45'} is not a legal
month.

@item
The MySQL server only performs basic checking on the validity of a date:
days @code{00-31}, months @code{00-12}, years @code{1000-9999}.
Any date not within this range will revert to @code{0000-00-00}.
Please note that this still allows you to store invalid dates such as
@code{2002-04-31}. It allows web applications to store data from a form
without further checking. To ensure a date is valid, perform a check in
your application.

@item
Year values specified as two digits are ambiguous, because the century is
unknown.  MySQL interprets 2-digit year values using the following
rules:

@itemize @minus
@item
Year values in the range @code{00-69} are converted to @code{2000-2069}.

@item
Year values in the range @code{70-99} are converted to @code{1970-1999}.
@end itemize
@end itemize


@node TIME, YEAR, DATETIME, Date and time types
@subsubsection The @code{TIME} Type

@tindex TIME

MySQL retrieves and displays @code{TIME} values in @code{'HH:MM:SS'}
format (or @code{'HHH:MM:SS'} format for large hours values).  @code{TIME}
values may range from @code{'-838:59:59'} to @code{'838:59:59'}.  The reason
the hours part may be so large is that the @code{TIME} type may be used not
only to represent a time of day (which must be less than 24 hours), but also
elapsed time or a time interval between two events (which may be much greater
than 24 hours, or even negative).

You can specify @code{TIME} values in a variety of formats:

@itemize @bullet
@item
As a string in @code{'D HH:MM:SS.fraction'} format.  (Note that
MySQL doesn't yet store the fraction for the time column.)  One
can also use one of the following ``relaxed'' syntax:

@code{HH:MM:SS.fraction}, @code{HH:MM:SS}, @code{HH:MM}, @code{D HH:MM:SS},
@code{D HH:MM}, @code{D HH} or @code{SS}.  Here @code{D} is days between 0-33.

@item
As a string with no delimiters in @code{'HHMMSS'} format, provided that
it makes sense as a time.  For example, @code{'101112'} is understood as
@code{'10:11:12'}, but @code{'109712'} is illegal (it has a nonsensical
minute part) and becomes @code{'00:00:00'}.

@item
As a number in @code{HHMMSS} format, provided that it makes sense as a time.
For example, @code{101112} is understood as @code{'10:11:12'}.  The following
alternative formats are also understood: @code{SS}, @code{MMSS},@code{HHMMSS},
@code{HHMMSS.fraction}.  Note that MySQL doesn't yet store the
fraction part.

@item
As the result of a function that returns a value that is acceptable
in a @code{TIME} context, such as @code{CURRENT_TIME}.
@end itemize

For @code{TIME} values specified as strings that include a time part
delimiter, it is not necessary to specify two digits for hours, minutes, or
seconds values that are less than @code{10}.  @code{'8:3:2'} is the same as
@code{'08:03:02'}.

Be careful about assigning ``short'' @code{TIME} values to a @code{TIME}
column. Without colons, MySQL interprets values using the
assumption that the rightmost digits represent seconds. (MySQL
interprets @code{TIME} values as elapsed time rather than as time of
day.) For example, you might think of @code{'1112'} and @code{1112} as
meaning @code{'11:12:00'} (12 minutes after 11 o'clock), but
MySQL interprets them as @code{'00:11:12'} (11 minutes, 12 seconds).
Similarly, @code{'12'} and @code{12} are interpreted as @code{'00:00:12'}.
@code{TIME} values with colons, by contrast, are always treated as
time of the day. That is @code{'11:12'} will mean @code{'11:12:00'},
not @code{'00:11:12'}.

Values that lie outside the @code{TIME} range
but are otherwise legal are clipped to the appropriate
endpoint of the range.  For example, @code{'-850:00:00'} and
@code{'850:00:00'} are converted to @code{'-838:59:59'} and
@code{'838:59:59'}.

Illegal @code{TIME} values are converted to @code{'00:00:00'}.  Note that
because @code{'00:00:00'} is itself a legal @code{TIME} value, there is no way
to tell, from a value of @code{'00:00:00'} stored in a table, whether the
original value was specified as @code{'00:00:00'} or whether it was illegal.


@node YEAR,  , TIME, Date and time types
@subsubsection The @code{YEAR} Type

@tindex YEAR

The @code{YEAR} type is a 1-byte type used for representing years.

MySQL retrieves and displays @code{YEAR} values in @code{YYYY}
format.  The range is @code{1901} to @code{2155}.

You can specify @code{YEAR} values in a variety of formats:

@itemize @bullet
@item
As a four-digit string in the range @code{'1901'} to @code{'2155'}.

@item
As a four-digit number in the range @code{1901} to @code{2155}.

@item
As a two-digit string in the range @code{'00'} to @code{'99'}.  Values in the
ranges @code{'00'} to @code{'69'} and @code{'70'} to @code{'99'} are
converted to @code{YEAR} values in the ranges @code{2000} to @code{2069} and
@code{1970} to @code{1999}.

@item
As a two-digit number in the range @code{1} to @code{99}.  Values in the
ranges @code{1} to @code{69} and @code{70} to @code{99} are converted to
@code{YEAR} values in the ranges @code{2001} to @code{2069} and @code{1970}
to @code{1999}.  Note that the range for two-digit numbers is slightly
different from the range for two-digit strings, because you cannot specify zero
directly as a number and have it be interpreted as @code{2000}.  You
@strong{must} specify it as a string @code{'0'} or @code{'00'} or it will be
interpreted as @code{0000}.

@item
As the result of a function that returns a value that is acceptable
in a @code{YEAR} context, such as @code{NOW()}.
@end itemize

Illegal @code{YEAR} values are converted to @code{0000}.


@node String types, Choosing types, Date and time types, Column types
@subsection String Types

@cindex types, strings
@cindex string types

@tindex CHAR
@tindex VARCHAR

@menu
* CHAR::                        The @code{CHAR} and @code{VARCHAR} types
* BLOB::                        The @code{BLOB} and @code{TEXT} types
* ENUM::                        The @code{ENUM} type
* SET::                         The @code{SET} type
@end menu


The string types are @code{CHAR}, @code{VARCHAR}, @code{BLOB}, @code{TEXT},
@code{ENUM}, and @code{SET}.  This section describes how these types work,
their storage requirements, and how to use them in your queries.


@node CHAR, BLOB, String types, String types
@subsubsection The @code{CHAR} and @code{VARCHAR} Types

The @code{CHAR} and @code{VARCHAR} types are similar, but differ in the
way they are stored and retrieved.

The length of a @code{CHAR} column is fixed to the length that you declare
when you create the table.  The length can be any value between 1 and 255.
(As of MySQL Version 3.23, the length of @code{CHAR} may be 0 to 255.)
When @code{CHAR} values are stored, they are right-padded with spaces to the
specified length.  When @code{CHAR} values are retrieved, trailing spaces are
removed.

Values in @code{VARCHAR} columns are variable-length strings.  You can
declare a @code{VARCHAR} column to be any length between 1 and 255, just as
for @code{CHAR} columns.  However, in contrast to @code{CHAR}, @code{VARCHAR}
values are stored using only as many characters as are needed, plus one byte
to record the length.  Values are not padded; instead, trailing spaces are
removed when values are stored.  (This space removal differs from the ANSI
SQL specification.)

If you assign a value to a @code{CHAR} or @code{VARCHAR} column that
exceeds the column's maximum length, the value is truncated to fit.

The following table illustrates the differences between the two types of columns
by showing the result of storing various string values into @code{CHAR(4)}
and @code{VARCHAR(4)} columns:

@c Need to use @(space) to make sure second column values retain spacing
@c in output for table below.

@multitable @columnfractions .15 .15 .20 .15 .20
@item @strong{Value}  @tab @code{CHAR(4)}     @tab @strong{Storage required} @tab @code{VARCHAR(4)} @tab @strong{Storage required}
@item @code{''}       @tab @code{'@ @ @ @ '} @tab 4 bytes @tab @code{''} @tab 1 byte
@item @code{'ab'}     @tab @code{'ab@ @ '}   @tab 4 bytes @tab @code{'ab'} @tab 3 bytes
@item @code{'abcd'}   @tab @code{'abcd'}     @tab 4 bytes @tab @code{'abcd'} @tab 5 bytes
@item @code{'abcdefgh'} @tab @code{'abcd'}     @tab 4 bytes @tab @code{'abcd'} @tab 5 bytes
@end multitable

The values retrieved from the @code{CHAR(4)} and @code{VARCHAR(4)} columns
will be the same in each case, because trailing spaces are removed from
@code{CHAR} columns upon retrieval.

Values in @code{CHAR} and @code{VARCHAR} columns are sorted and compared
in case-insensitive fashion, unless the @code{BINARY} attribute was
specified when the table was created.  The @code{BINARY} attribute means
that column values are sorted and compared in case-sensitive fashion
according to the ASCII order of the machine where the MySQL
server is running. @code{BINARY} doesn't affect how the column is stored
or retrieved.

The @code{BINARY} attribute is sticky.  This means that if a column marked
@code{BINARY} is used in an expression, the whole expression is compared as a
@code{BINARY} value.

MySQL may silently change the type of a @code{CHAR} or @code{VARCHAR}
column at table creation time.
@xref{Silent column changes}.


@node BLOB, ENUM, CHAR, String types
@subsubsection The @code{BLOB} and @code{TEXT} Types

@tindex BLOB
@tindex TEXT

A @code{BLOB} is a binary large object that can hold a variable amount of
data.  The four @code{BLOB} types @code{TINYBLOB}, @code{BLOB},
@code{MEDIUMBLOB}, and @code{LONGBLOB} differ only in the maximum length of
the values they can hold.
@xref{Storage requirements}.

The four @code{TEXT} types @code{TINYTEXT}, @code{TEXT}, @code{MEDIUMTEXT},
and @code{LONGTEXT} correspond to the four @code{BLOB} types and have the
same maximum lengths and storage requirements.  The only difference between
@code{BLOB} and @code{TEXT} types is that sorting and comparison is performed
in case-sensitive fashion for @code{BLOB} values and case-insensitive fashion
for @code{TEXT} values.  In other words, a @code{TEXT} is a case-insensitive
@code{BLOB}.

If you assign a value to a @code{BLOB} or @code{TEXT} column that exceeds
the column type's maximum length, the value is truncated to fit.

In most respects, you can regard a @code{TEXT} column as a @code{VARCHAR}
column that can be as big as you like.  Similarly, you can regard a
@code{BLOB} column as a @code{VARCHAR BINARY} column.  The differences are:

@itemize @bullet
@item
You can have indexes on @code{BLOB} and @code{TEXT} columns with
MySQL Version 3.23.2 and newer. Older versions of
MySQL did not support this.

@item
There is no trailing-space removal for @code{BLOB} and @code{TEXT} columns
when values are stored, as there is for @code{VARCHAR} columns.

@item
@cindex default values, @code{BLOB} and @code{TEXT} columns
@cindex @code{BLOB} columns, default values
@cindex @code{TEXT} columns, default values
@code{BLOB} and @code{TEXT} columns cannot have @code{DEFAULT} values.
@end itemize

@code{MyODBC} defines @code{BLOB} values as @code{LONGVARBINARY} and
@code{TEXT} values as @code{LONGVARCHAR}.

Because @code{BLOB} and @code{TEXT} values may be extremely long, you
may run up against some constraints when using them:

@itemize @bullet
@item
If you want to use @code{GROUP BY} or @code{ORDER BY} on a @code{BLOB} or
@code{TEXT} column, you must convert the column value into a fixed-length
object. The standard way to do this is with the @code{SUBSTRING}
function.  For example:

@example
mysql> SELECT comment FROM tbl_name,SUBSTRING(comment,20) AS substr
    ->                 ORDER BY substr;
@end example

If you don't do this, only the first @code{max_sort_length} bytes of the
column are used when sorting.  The default value of @code{max_sort_length} is
1024; this value can be changed using the @code{-O} option when starting the
@code{mysqld} server. You can group on an expression involving @code{BLOB} or
@code{TEXT} values by specifying the column position or by using an alias:

@example
mysql> SELECT id,SUBSTRING(blob_col,1,100) FROM tbl_name GROUP BY 2;
mysql> SELECT id,SUBSTRING(blob_col,1,100) AS b FROM tbl_name GROUP BY b;
@end example

@item
The maximum size of a @code{BLOB} or @code{TEXT} object is determined by its
type, but the largest value you can actually transmit between the client and
server is determined by the amount of available memory and the size of the
communications buffers.  You can change the message buffer size, but you must
do so on both the server and client ends. @xref{Server parameters}.
@end itemize

Note that each @code{BLOB} or @code{TEXT} value is represented
internally by a separately allocated object. This is in contrast to all
other column types, for which storage is allocated once per column when
the table is opened.


@node ENUM, SET, BLOB, String types
@subsubsection The @code{ENUM} Type

@tindex ENUM

An @code{ENUM} is a string object whose value normally is chosen from a list
of allowed values that are enumerated explicitly in the column specification
at table creation time.

The value may also be the empty string (@code{""}) or @code{NULL} under
certain circumstances:

@itemize @bullet
@item
If you insert an invalid value into an @code{ENUM} (that is, a string not
present in the list of allowed values), the empty string is inserted
instead as a special error value. This string can be distinguished from a
'normal' empty string by the fact that this string has the numerical value
0. More about this later.

@item
If an @code{ENUM} is declared @code{NULL}, @code{NULL} is also a legal value
for the column, and the default value is @code{NULL}.  If an @code{ENUM} is
declared @code{NOT NULL}, the default value is the first element of the
list of allowed values.
@end itemize

Each enumeration value has an index:

@itemize @bullet
@item
Values from the list of allowable elements in the column specification are
numbered beginning with 1.

@item
The index value of the empty string error value is 0.  This means that you
can use the following @code{SELECT} statement to find rows into which invalid
@code{ENUM} values were assigned:

@example
mysql> SELECT * FROM tbl_name WHERE enum_col=0;
@end example

@item
The index of the @code{NULL} value is @code{NULL}.
@end itemize

For example, a column specified as @code{ENUM("one", "two", "three")} can
have any of the values shown here.  The index of each value is also shown:

@multitable @columnfractions .10 .10
@item @strong{Value} @tab @strong{Index}
@item @code{NULL}    @tab @code{NULL}
@item @code{""}      @tab 0
@item @code{"one"}   @tab 1
@item @code{"two"}   @tab 2
@item @code{"three"} @tab 3
@end multitable

An enumeration can have a maximum of 65535 elements.

Starting from 3.23.51 trailing spaces are automatically deleted from
@code{ENUM} values when the table is created.

Lettercase is irrelevant when you assign values to an @code{ENUM} column.
However, values retrieved from the column later have lettercase matching the
values that were used to specify the allowable values at table creation time.

If you retrieve an @code{ENUM} in a numeric context, the column value's
index is returned.  For example, you can retrieve numeric values from
an @code{ENUM} column like this:

@example
mysql> SELECT enum_col+0 FROM tbl_name;
@end example

If you store a number into an @code{ENUM}, the number is treated as an
index, and the value stored is the enumeration member with that index.
(However, this will not work with @code{LOAD DATA}, which treats all
input as strings.)
It's not advisable to store numbers in an @code{ENUM} string because
it will make things confusing.

@code{ENUM} values are sorted according to the order in which the enumeration
members were listed in the column specification.  (In other words,
@code{ENUM} values are sorted according to their index numbers.)  For
example, @code{"a"} sorts before @code{"b"} for @code{ENUM("a", "b")}, but
@code{"b"} sorts before @code{"a"} for @code{ENUM("b", "a")}.  The empty
string sorts before non-empty strings, and @code{NULL} values sort before
all other enumeration values.

If you want to get all possible values for an @code{ENUM} column, you should
use: @code{SHOW COLUMNS FROM table_name LIKE enum_column_name} and parse
the @code{ENUM} definition in the second column.


@node SET,  , ENUM, String types
@subsubsection The @code{SET} Type

@tindex SET

A @code{SET} is a string object that can have zero or more values, each of
which must be chosen from a list of allowed values specified when the table
is created.  @code{SET} column values that consist of multiple set members
are specified with members separated by commas (@samp{,}).  A consequence of
this is that @code{SET} member values cannot themselves contain commas.

For example, a column specified as @code{SET("one", "two") NOT NULL} can have
any of these values:

@example
""
"one"
"two"
"one,two"
@end example


A @code{SET} can have a maximum of 64 different members.

Starting from 3.23.51 trailing spaces are automatically deleted from
@code{SET} values when the table is created.

MySQL stores @code{SET} values numerically, with the low-order bit
of the stored value corresponding to the first set member.  If you retrieve a
@code{SET} value in a numeric context, the value retrieved has bits set
corresponding to the set members that make up the column value.  For example,
you can retrieve numeric values from a @code{SET} column like this:

@example
mysql> SELECT set_col+0 FROM tbl_name;
@end example

If a number is stored into a @code{SET} column, the bits that
are set in the binary representation of the number determine the
set members in the column value.  Suppose a column is specified as
@code{SET("a","b","c","d")}.  Then the members have the following bit
values:

@multitable @columnfractions .15 .20 .20
@item @code{SET} @strong{member} @tab @strong{Decimal value} @tab @strong{Binary value}
@item @code{a} @tab @code{1} @tab @code{0001}
@item @code{b} @tab @code{2} @tab @code{0010}
@item @code{c} @tab @code{4} @tab @code{0100}
@item @code{d} @tab @code{8} @tab @code{1000}
@end multitable

If you assign a value of @code{9} to this column, that is @code{1001} in
binary, so the first and fourth @code{SET} value members @code{"a"} and
@code{"d"} are selected and the resulting value is @code{"a,d"}.

For a value containing more than one @code{SET} element, it does not matter
what order the elements are listed in when you insert the value.  It also
does not matter how many times a given element is listed in the value.
When the value is retrieved later, each element in the value will appear
once, with elements listed according to the order in which they were
specified at table creation time.  For example, if a column is specified as
@code{SET("a","b","c","d")}, then @code{"a,d"}, @code{"d,a"}, and
@code{"d,a,a,d,d"} will all appear as @code{"a,d"} when retrieved.

If you set a @code{SET} column to an unsupported value, the value will
be ignored.

@code{SET} values are sorted numerically.  @code{NULL} values sort before
non-@code{NULL} @code{SET} values.

Normally, you perform a @code{SELECT} on a @code{SET} column using
the @code{LIKE} operator or the @code{FIND_IN_SET()} function:

@example
mysql> SELECT * FROM tbl_name WHERE set_col LIKE '%value%';
mysql> SELECT * FROM tbl_name WHERE FIND_IN_SET('value',set_col)>0;
@end example

But the following will also work:

@example
mysql> SELECT * FROM tbl_name WHERE set_col = 'val1,val2';
mysql> SELECT * FROM tbl_name WHERE set_col & 1;
@end example

The first of these statements looks for an exact match.  The second looks
for values containing the first set member.

If you want to get all possible values for a @code{SET} column, you should
use: @code{SHOW COLUMNS FROM table_name LIKE set_column_name} and parse
the @code{SET} definition in the second column.


@node Choosing types, Other-vendor column types, String types, Column types
@subsection Choosing the Right Type for a Column

@cindex types, columns
@cindex choosing types

For the most efficient use of storage, try to use the most precise type in
all cases. For example, if an integer column will be used for values in the
range between @code{1} and @code{99999}, @code{MEDIUMINT UNSIGNED} is the
best type.

Accurate representation of monetary values is a common problem. In
MySQL, you should use the @code{DECIMAL} type. This is stored as
a string, so no loss of accuracy should occur. If accuracy is not
too important, the @code{DOUBLE} type may also be good enough.

For high precision, you can always convert to a fixed-point type stored
in a @code{BIGINT}. This allows you to do all calculations with integers
and convert results back to floating-point values only when necessary.


@node Other-vendor column types, Storage requirements, Choosing types, Column types
@subsection Using Column Types from Other Database Engines

@cindex types, portability
@cindex portability, types
@cindex columns, other types

To make it easier to use code written for SQL implementations from other
vendors, MySQL maps column types as shown in the following table.  These
mappings make it easier to move table definitions from other database engines
to MySQL:

@multitable @columnfractions .25 .30
@item @strong{Other vendor type}        @tab @strong{MySQL type}
@item @code{BINARY(NUM)}		@tab @code{CHAR(NUM) BINARY}
@item @code{CHAR VARYING(NUM)}		@tab @code{VARCHAR(NUM)}
@item @code{FLOAT4}			@tab @code{FLOAT}
@item @code{FLOAT8}			@tab @code{DOUBLE}
@item @code{INT1}			@tab @code{TINYINT}
@item @code{INT2}			@tab @code{SMALLINT}
@item @code{INT3}			@tab @code{MEDIUMINT}
@item @code{INT4}			@tab @code{INT}
@item @code{INT8}			@tab @code{BIGINT}
@item @code{LONG VARBINARY}		@tab @code{MEDIUMBLOB}
@item @code{LONG VARCHAR}		@tab @code{MEDIUMTEXT}
@item @code{MIDDLEINT}			@tab @code{MEDIUMINT}
@item @code{VARBINARY(NUM)}		@tab @code{VARCHAR(NUM) BINARY}
@end multitable

Column type mapping occurs at table creation time.  If you create a table
with types used by other vendors and then issue a @code{DESCRIBE tbl_name}
statement, MySQL reports the table structure using the equivalent
MySQL types.


@node Storage requirements,  , Other-vendor column types, Column types
@subsection Column Type Storage Requirements

@cindex storage requirements, column type
@cindex columns, storage requirements

The storage requirements for each of the column types supported by
MySQL are listed by category.

@cindex numeric types
@cindex types, numeric

@subsubheading Storage requirements for numeric types

@multitable @columnfractions .35 .65
@item @strong{Column type} @tab @strong{Storage required}
@item @code{TINYINT}       @tab 1 byte
@item @code{SMALLINT}      @tab 2 bytes
@item @code{MEDIUMINT}     @tab 3 bytes
@item @code{INT}           @tab 4 bytes
@item @code{INTEGER}       @tab 4 bytes
@item @code{BIGINT}        @tab 8 bytes
@item @code{FLOAT(X)}      @tab 4 if X <= 24 or 8 if 25 <= X <= 53
@item @code{FLOAT}         @tab 4 bytes
@item @code{DOUBLE}        @tab 8 bytes
@item @code{DOUBLE PRECISION} @tab 8 bytes
@item @code{REAL}         @tab 8 bytes
@item @code{DECIMAL(M,D)} @tab @code{M+2} bytes if D > 0, @code{M+1} bytes if D = 0 (@code{D}+2, if @code{M < D})
@item @code{NUMERIC(M,D)} @tab @code{M+2} bytes if D > 0, @code{M+1} bytes if D = 0 (@code{D}+2, if @code{M < D})
@end multitable

@cindex date types
@cindex time types
@cindex types, date
@cindex types, time

@subsubheading Storage requirements for date and time types

@multitable @columnfractions .35 .65
@item @strong{Column type} @tab @strong{Storage required}
@item @code{DATE}          @tab 3 bytes
@item @code{DATETIME}      @tab 8 bytes
@item @code{TIMESTAMP}     @tab 4 bytes
@item @code{TIME}          @tab 3 bytes
@item @code{YEAR}          @tab 1 byte
@end multitable

@subsubheading Storage requirements for string types

@multitable @columnfractions .35 .65
@item @strong{Column type} @tab @strong{Storage required}
@item @code{CHAR(M)} @tab @code{M} bytes, @code{1 <= M <= 255}
@item @code{VARCHAR(M)} @tab @code{L}+1 bytes, where @code{L <= M} and
@code{1 <= M <= 255}
@item @code{TINYBLOB}, @code{TINYTEXT} @tab @code{L}+1 bytes,
where @code{L} < 2^8
@item @code{BLOB}, @code{TEXT} @tab @code{L}+2 bytes,
where @code{L} < 2^16
@item @code{MEDIUMBLOB}, @code{MEDIUMTEXT} @tab @code{L}+3 bytes,
where @code{L} < 2^24
@item @code{LONGBLOB}, @code{LONGTEXT} @tab @code{L}+4 bytes,
where @code{L} < 2^32
@item @code{ENUM('value1','value2',...)} @tab 1 or 2 bytes, depending on
the number of enumeration values (65535 values maximum)
@item @code{SET('value1','value2',...)} @tab 1, 2, 3, 4 or 8 bytes, depending
on the number of set members (64 members maximum)
@end multitable

@cindex @code{BLOB}, size
@cindex @code{TEXT}, size
@cindex @code{VARCHAR}, size
@code{VARCHAR} and the @code{BLOB} and @code{TEXT} types are variable-length
types, for which the storage requirements depend on the actual length of
column values (represented by @code{L} in the preceding table), rather than
on the type's maximum possible size.  For example, a @code{VARCHAR(10)}
column can hold a string with a maximum length of 10 characters.  The actual
storage required is the length of the string (@code{L}), plus 1 byte to
record the length of the string.  For the string @code{'abcd'}, @code{L} is 4
and the storage requirement is 5 bytes.

The @code{BLOB} and @code{TEXT} types require 1, 2, 3, or 4 bytes to record
the length of the column value, depending on the maximum possible length of
the type.  @xref{BLOB}.

If a table includes any variable-length column types, the record format will
also be variable-length.  Note that when a table is created, MySQL
may, under certain conditions, change a column from a variable-length type to a
fixed-length type, or vice-versa.  @xref{Silent column changes}.

@cindex ENUM, size
The size of an @code{ENUM} object is determined by the number of
different enumeration values.  One byte is used for enumerations with up
to 255 possible values.  Two bytes are used for enumerations with up to
65535 values. @xref{ENUM}.

@cindex SET, size
The size of a @code{SET} object is determined by the number of different
set members.  If the set size is @code{N}, the object occupies @code{(N+7)/8}
bytes, rounded up to 1, 2, 3, 4, or 8 bytes.  A @code{SET} can have a maximum
of 64 members. @xref{SET}.


@node Functions, Data Manipulation, Column types, Reference
@section Functions for Use in @code{SELECT} and @code{WHERE} Clauses

@cindex functions for @code{SELECT} and @code{WHERE} clauses

A @code{select_expression} or @code{where_definition} in a SQL statement
can consist of any expression using the functions described below.

An expression that contains @code{NULL} always produces a @code{NULL} value
unless otherwise indicated in the documentation for the operators and
functions involved in the expression.

@strong{Note}: there must be no whitespace between a function name and the
parentheses following it. This helps the MySQL parser distinguish
between function calls and references to tables or columns that happen to
have the same name as a function.  Spaces around arguments are permitted,
though.

You can force MySQL to accept spaces after the function name by
starting @code{mysqld} with @code{--ansi} or using the
@code{CLIENT_IGNORE_SPACE} to @code{mysql_connect()}, but in this case all
function names will become reserved words. @xref{ANSI mode}.

@need 2000
For the sake of brevity, examples display the output from the @code{mysql}
program in abbreviated form.  So this:

@example
mysql> SELECT MOD(29,9);
1 rows in set (0.00 sec)

+-----------+
| mod(29,9) |
+-----------+
|         2 |
+-----------+
@end example

is displayed like this:

@example
mysql> SELECT MOD(29,9);
        -> 2
@end example

@menu
* Non-typed Operators::         Non-Type-Specific Operators and Functions
* String functions::            String functions
* Numeric Functions::           Numeric Functions
* Date and time functions::     Date and time functions
* Cast Functions::              Cast Functions
* Other Functions::             Other Functions
* Group by functions::          Functions for Use with @code{GROUP BY} Clauses
@end menu


@node Non-typed Operators, String functions, Functions, Functions
@subsection Non-Type-Specific Operators and Functions

@menu
* Parentheses::                 Parentheses
* Comparison Operators::        Comparison Operators
* Logical Operators::           Logical Operators
* Control flow functions::      Control Flow Functions
@end menu


@node Parentheses, Comparison Operators, Non-typed Operators, Non-typed Operators
@subsubsection Parentheses

@findex () (parentheses)
@findex parentheses ( and )

@cindex functions, grouping
@cindex grouping, expressions

@example
( ... )
@end example

Use parentheses to force the order of evaluation in an expression.  For
example:

@example
mysql> SELECT 1+2*3;
        -> 7
mysql> SELECT (1+2)*3;
        -> 9
@end example



@node Comparison Operators, Logical Operators, Parentheses, Non-typed Operators
@subsubsection Comparison Operators

@findex comparison operators

@cindex casts
@cindex type conversions

Comparison operations result in a value of @code{1} (TRUE), @code{0} (FALSE),
or @code{NULL}. These functions work for both numbers and strings.  Strings
are automatically converted to numbers and numbers to strings as needed (as
in Perl).

MySQL performs comparisons using the following
rules:

@itemize @bullet
@item
If one or both arguments are @code{NULL}, the result of the comparison
is @code{NULL}, except for the @code{<=>} operator.

@item
If both arguments in a comparison operation are strings, they are compared as
strings.

@item
If both arguments are integers, they are compared as integers.

@item
Hexadecimal values are treated as binary strings if not compared to a number.

@item
@cindex ODBC compatibility
@cindex compatibility, with ODBC
If one of the arguments is a @code{TIMESTAMP} or @code{DATETIME} column and
the other argument is a constant, the constant is converted
to a timestamp before the comparison is performed. This is done to be more
ODBC-friendly.

@item
In all other cases, the arguments are compared as floating-point (real)
numbers.
@end itemize

By default, string comparisons are done in case-independent fashion using the
current character set (ISO-8859-1 Latin1 by default, which also works
excellently for English).

The following examples illustrate conversion of strings to numbers for comparison
operations:

@example
mysql> SELECT 1 > '6x';
         -> 0
mysql> SELECT 7 > '6x';
         -> 1
mysql> SELECT 0 > 'x6';
         -> 0
mysql> SELECT 0 = 'x6';
         -> 1
@end example

@table @code
@findex = (equal)
@findex equal (=)
@item =
Equal:
@example
mysql> SELECT 1 = 0;
        -> 0
mysql> SELECT '0' = 0;
        -> 1
mysql> SELECT '0.0' = 0;
        -> 1
mysql> SELECT '0.01' = 0;
        -> 0
mysql> SELECT '.01' = 0.01;
        -> 1
@end example

@findex <> (not equal)
@findex not equal (<>)
@findex != (not equal)
@findex not equal (!=)
@item <>
@itemx !=
Not equal:
@example
mysql> SELECT '.01' <> '0.01';
        -> 1
mysql> SELECT .01 <> '0.01';
        -> 0
mysql> SELECT 'zapp' <> 'zappp';
        -> 1
@end example

@findex <= (less than or equal)
@findex less than or equal (<=)
@item <=
Less than or equal:
@example
mysql> SELECT 0.1 <= 2;
        -> 1
@end example

@findex < (less than)
@findex less than (<)
@item <
Less than:
@example
mysql> SELECT 2 < 2;
        -> 0
@end example

@findex >= (greater than or equal)
@findex greater than or equal (>=)
@item >=
Greater than or equal:
@example
mysql> SELECT 2 >= 2;
        -> 1
@end example

@findex > (greater than)
@findex greater than (>)
@item >
Greater than:
@example
mysql> SELECT 2 > 2;
        -> 0
@end example

@cindex @code{NULL}, testing for null
@findex <=> (Equal to)
@item <=>
NULL safe equal:
@example
mysql> SELECT 1 <=> 1, NULL <=> NULL, 1 <=> NULL;
        -> 1 1 0
@end example

@cindex @code{NULL}, testing for null
@findex IS NULL
@findex IS NOT NULL
@item IS NULL
@itemx IS NOT NULL
Test whether a value is or is not @code{NULL}:
@example
mysql> SELECT 1 IS NULL, 0 IS NULL, NULL IS NULL;
        -> 0 0 1
mysql> SELECT 1 IS NOT NULL, 0 IS NOT NULL, NULL IS NOT NULL;
        -> 1 1 0
@end example

@cindex ODBC compatibility
@cindex compatibility, with ODBC
To be able to work good with other programs, MySQL supports the following
extra features when using @code{IS NULL}:

@itemize @bullet
@item
You can find the last inserted row with:

@example
SELECT * FROM tbl_name WHERE auto_col IS NULL
@end example

This can be disabled by setting @code{SQL_AUTO_IS_NULL=0}. @xref{SET OPTION}.
@item
For @code{NOT NULL} @code{DATE} and @code{DATETIME} columns you can find
the special date @code{0000-00-00} by using:

@example
SELECT * FROM tbl_name WHERE date_column IS NULL
@end example

This is needed to get some ODBC applications to work (as ODBC doesn't
support a @code{0000-00-00} date)
@end itemize

@findex BETWEEN ... AND
@item expr BETWEEN min AND max
If @code{expr} is greater than or equal to @code{min} and @code{expr} is
less than or equal to @code{max}, @code{BETWEEN} returns @code{1},
otherwise it returns @code{0}.  This is equivalent to the expression
@code{(min <= expr AND expr <= max)} if all the arguments are of the
same type.  The first argument (@code{expr}) determines how the
comparison is performed as follows:

@itemize @bullet
@item
If @code{expr} is a @code{TIMESTAMP}, @code{DATE}, or @code{DATETIME}
column, @code{MIN()} and @code{MAX()} are formatted to the same format if
they are constants.
@item
If @code{expr} is a case-insensitive string expression, a case-insensitive
string comparison is done.
@item
If @code{expr} is a case-sensitive string expression, a case-sensitive
string comparison is done.
@item
If @code{expr} is an integer expression, an integer comparison is done.
@item
Otherwise, a floating-point (real) comparison is done.
@end itemize

@example
mysql> SELECT 1 BETWEEN 2 AND 3;
        -> 0
mysql> SELECT 'b' BETWEEN 'a' AND 'c';
        -> 1
mysql> SELECT 2 BETWEEN 2 AND '3';
        -> 1
mysql> SELECT 2 BETWEEN 2 AND 'x-3';
        -> 0
@end example

@findex NOT BETWEEN
@item expr NOT BETWEEN min AND max
Same as @code{NOT (expr BETWEEN min AND max)}.

@findex IN
@item expr IN (value,...)
Returns @code{1} if @code{expr} is any of the values in the @code{IN} list,
else returns @code{0}.  If all values are constants, then all values are
evaluated according to the type of @code{expr} and sorted. The search for the
item is then done using a binary search. This means @code{IN} is very quick
if the @code{IN} value list consists entirely of constants.  If @code{expr}
is a case-sensitive string expression, the string comparison is performed in
case-sensitive fashion:

@example
mysql> SELECT 2 IN (0,3,5,'wefwf');
        -> 0
mysql> SELECT 'wefwf' IN (0,3,5,'wefwf');
        -> 1
@end example

@findex NOT IN
@item expr NOT IN (value,...)
Same as @code{NOT (expr IN (value,...))}.

@findex ISNULL()
@item ISNULL(expr)
If @code{expr} is @code{NULL}, @code{ISNULL()} returns @code{1}, otherwise
it returns @code{0}:
@example
mysql> SELECT ISNULL(1+1);
        -> 0
mysql> SELECT ISNULL(1/0);
        -> 1
@end example

Note that a comparison of @code{NULL} values using @code{=} will always be
false!

@cindex @code{NULL}, testing for null
@findex COALESCE()
@item COALESCE(list)
Returns first non-@code{NULL} element in list:

@example
mysql> SELECT COALESCE(NULL,1);
        -> 1
mysql> SELECT COALESCE(NULL,NULL,NULL);
        -> NULL
@end example

@findex INTERVAL()
@item INTERVAL(N,N1,N2,N3,...)
Returns @code{0} if @code{N} < @code{N1}, @code{1} if @code{N} < @code{N2}
and so on. All arguments are treated as integers.  It is required that
@code{N1} < @code{N2} < @code{N3} < @code{...} < @code{Nn} for this function
to work correctly. This is because a binary search is used (very fast):

@example
mysql> SELECT INTERVAL(23, 1, 15, 17, 30, 44, 200);
        -> 3
mysql> SELECT INTERVAL(10, 1, 10, 100, 1000);
        -> 2
mysql> SELECT INTERVAL(22, 23, 30, 44, 200);
        -> 0
@end example
@end table

If you are comparing case-sensitive string with any of the standard
operators (@code{=}, @code{<>}..., but not @code{LIKE}) end space will
be ignored.

@example
mysql> SELECT "a" ="A ";
        -> 1
@end example


@node Logical Operators, Control flow functions, Comparison Operators, Non-typed Operators
@subsubsection Logical Operators

@findex Logical operators
@findex Operators, logical

All logical operators evaluate to @code{1} (TRUE), @code{0} (FALSE) or
@code{NULL} (unknown, which is in most cases the same as FALSE):

@table @code
@findex NOT, logical
@findex ! (logical NOT)
@item NOT
@itemx !
Logical NOT. Evaluates to @code{1} if the operand is @code{0}, otherwise
evaluates to @code{0}.
Exception: @code{NOT NULL} evaluates to @code{NULL}:
@example
mysql> SELECT NOT 1;
        -> 0
mysql> SELECT NOT NULL;
        -> NULL
mysql> SELECT ! (1+1);
        -> 0
mysql> SELECT ! 1+1;
        -> 1
@end example
The last example produces @code{1} because the expression evaluates
the same way as @code{(!1)+1}.

@findex OR, logical
@findex || (logical OR)
@item OR
@itemx ||
Logical OR. Evaluates to @code{1} if either operand is not @code{0} and not
@code{NULL}:
@example
mysql> SELECT 1 || 0;
        -> 1
mysql> SELECT 0 || 0;
        -> 0
mysql> SELECT 1 || NULL;
        -> 1

@end example

@findex AND, logical
@findex && (logical AND)
@item AND
@itemx &&
Logical AND. For non-@code{NULL} operands, evaluates to @code{1} if both
operands are non-zero and to @code{0} otherwise.
Produces @code{NULL} if either operand is @code{NULL}:
@example
mysql> SELECT 1 && 1;
        -> 1
mysql> SELECT 1 && 0;
        -> 0
mysql> SELECT 1 && NULL;
        -> NULL
@end example
@end table


@node Control flow functions,  , Logical Operators, Non-typed Operators
@subsubsection Control Flow Functions

@findex control flow functions
@findex functions, control flow

@table @code
@cindex @code{NULL}, testing for null
@findex IFNULL()
@item IFNULL(expr1,expr2)
If @code{expr1} is not @code{NULL}, @code{IFNULL()} returns @code{expr1},
else it returns @code{expr2}.  @code{IFNULL()} returns a numeric or string
value, depending on the context in which it is used:
@example
mysql> SELECT IFNULL(1,0);
        -> 1
mysql> SELECT IFNULL(NULL,10);
        -> 10
mysql> SELECT IFNULL(1/0,10);
        -> 10
mysql> SELECT IFNULL(1/0,'yes');
        -> 'yes'
@end example

@findex NULLIF()
@item NULLIF(expr1,expr2)
If @code{expr1 = expr2} is true, return @code{NULL} else return @code{expr1}.
This is the same as @code{CASE WHEN x = y THEN NULL ELSE x END}:
@example
mysql> SELECT NULLIF(1,1);
        -> NULL
mysql> SELECT NULLIF(1,2);
        -> 1
@end example

Note that @code{expr1} is evaluated twice in MySQL if the arguments
are equal.

@findex IF()
@item IF(expr1,expr2,expr3)
If @code{expr1} is TRUE (@code{expr1 <> 0} and @code{expr1 <> NULL}) then
@code{IF()} returns @code{expr2}, else it returns @code{expr3}.
@code{IF()} returns a numeric or string value, depending on the context
in which it is used:

@example
mysql> SELECT IF(1>2,2,3);
        -> 3
mysql> SELECT IF(1<2,'yes','no');
        -> 'yes'
mysql> SELECT IF(STRCMP('test','test1'),'no','yes');
        -> 'no'
@end example

@code{expr1} is evaluated as an integer value, which means that if you are
testing floating-point or string values, you should do so using a comparison
operation:

@example
mysql> SELECT IF(0.1,1,0);
        -> 0
mysql> SELECT IF(0.1<>0,1,0);
        -> 1
@end example

In the first case above, @code{IF(0.1)} returns @code{0} because @code{0.1}
is converted to an integer value, resulting in a test of @code{IF(0)}.  This
may not be what you expect.  In the second case, the comparison tests the
original floating-point value to see whether it is non-zero.  The result
of the comparison is used as an integer.

The default return type of @code{IF()} (which may matter when it is
stored into a temporary table) is calculated in MySQL Version
3.23 as follows:

@multitable @columnfractions .50 .20
@item @strong{Expression} @tab @strong{Return value}
@item expr2 or expr3 returns string @tab string
@item expr2 or expr3 returns a floating-point value @tab floating-point
@item expr2 or expr3 returns an integer @tab integer
@end multitable

If expr2 and expr3 are strings, then the result is case-sensitive if
both strings are case-sensitive. (Starting from 3.23.51)

@findex CASE
@item CASE value WHEN [compare-value] THEN result [WHEN [compare-value] THEN result ...] [ELSE result] END
@item CASE WHEN [condition] THEN result [WHEN [condition] THEN result ...] [ELSE result] END

The first version returns the @code{result} where
@code{value=compare-value}. The second version returns the result for
the first condition, which is true. If there was no matching result
value, then the result after @code{ELSE} is returned. If there is no
@code{ELSE} part then @code{NULL} is returned:

@example
mysql> SELECT CASE 1 WHEN 1 THEN "one"
           WHEN 2 THEN "two" ELSE "more" END;
       -> "one"
mysql> SELECT CASE WHEN 1>0 THEN "true" ELSE "false" END;
       -> "true"
mysql> SELECT CASE BINARY "B" WHEN "a" THEN 1 WHEN "b" THEN 2 END;
       -> NULL
@end example
@end table

The type of the return value (@code{INTEGER}, @code{DOUBLE} or
@code{STRING}) is the same as the type of the first returned value (the
expression after the first @code{THEN}).


@node String functions, Numeric Functions, Non-typed Operators, Functions
@subsection String Functions

@findex string functions
@findex functions, string

String-valued functions return @code{NULL} if the length of the result would
be greater than the @code{max_allowed_packet} server parameter.  @xref{Server
parameters}.

For functions that operate on string positions,
the first position is numbered 1.

@table @code
@findex ASCII()
@item ASCII(str)
Returns the ASCII code value of the leftmost character of the string
@code{str}. Returns @code{0} if @code{str} is the empty string.  Returns
@code{NULL} if @code{str} is @code{NULL}:

@example
mysql> SELECT ASCII('2');
        -> 50
mysql> SELECT ASCII(2);
        -> 50
mysql> SELECT ASCII('dx');
        -> 100
@end example

See also the @code{ORD()} function.

@findex ORD()
@item ORD(str)
If the leftmost character of the string @code{str} is a multi-byte character,
returns the code for that character, calculated from the ASCII code values
of its constituent characters using this formula:
@code{((first byte ASCII code)*256+(second byte ASCII code))[*256+third byte ASCII code...]}.
If the leftmost character is not a multi-byte character, returns the same
value that the @code{ASCII()} function does:

@example
mysql> SELECT ORD('2');
        -> 50
@end example

@findex CONV()
@item CONV(N,from_base,to_base)
Converts numbers between different number bases.  Returns a string
representation of the number @code{N}, converted from base @code{from_base}
to base @code{to_base}.  Returns @code{NULL} if any argument is @code{NULL}.
The argument @code{N} is interpreted as an integer, but may be specified as
an integer or a string.  The minimum base is @code{2} and the maximum base is
@code{36}.  If @code{to_base} is a negative number, @code{N} is regarded as a
signed number.  Otherwise, @code{N} is treated as unsigned.  @code{CONV} works
with 64-bit precision:

@example
mysql> SELECT CONV("a",16,2);
        -> '1010'
mysql> SELECT CONV("6E",18,8);
        -> '172'
mysql> SELECT CONV(-17,10,-18);
        -> '-H'
mysql> SELECT CONV(10+"10"+'10'+0xa,10,10);
        -> '40'
@end example

@findex BIN()
@item BIN(N)
Returns a string representation of the binary value of @code{N}, where
@code{N} is a longlong (@code{BIGINT}) number.  This is equivalent to
@code{CONV(N,10,2)}.  Returns @code{NULL} if @code{N} is @code{NULL}:

@example
mysql> SELECT BIN(12);
        -> '1100'
@end example

@findex OCT()
@item OCT(N)
Returns a string representation of the octal value of @code{N}, where
@code{N} is a longlong number.  This is equivalent to @code{CONV(N,10,8)}.
Returns @code{NULL} if @code{N} is @code{NULL}:

@example
mysql> SELECT OCT(12);
        -> '14'
@end example

@findex HEX()
@item HEX(N_or_S)

If N_OR_S is a number, returns a string representation of the hexadecimal
value of @code{N}, where @code{N} is a longlong (@code{BIGINT}) number.
This is equivalent to @code{CONV(N,10,16)}.

If N_OR_S is a string, returns a hexadecimal string of N_OR_S where each
character in N_OR_S is converted to 2 hexadecimal digits. This is the
invers of the @code{0xff} strings.

@example
mysql> SELECT HEX(255);
        -> 'FF'
mysql> SELECT HEX("abc");
        -> 616263
mysql> SELECT 0x616263;
        -> "abc"
@end example

@findex CHAR()
@item CHAR(N,...)
@code{CHAR()} interprets the arguments as integers and returns a string
consisting of the characters given by the ASCII code values of those
integers. @code{NULL} values are skipped:

@example
mysql> SELECT CHAR(77,121,83,81,'76');
        -> 'MySQL'
mysql> SELECT CHAR(77,77.3,'77.3');
        -> 'MMM'
@end example

@findex CONCAT()
@item CONCAT(str1,str2,...)
Returns the string that results from concatenating the arguments.  Returns
@code{NULL} if any argument is @code{NULL}.  May have more than 2 arguments.
A numeric argument is converted to the equivalent string form:

@example
mysql> SELECT CONCAT('My', 'S', 'QL');
        -> 'MySQL'
mysql> SELECT CONCAT('My', NULL, 'QL');
        -> NULL
mysql> SELECT CONCAT(14.3);
        -> '14.3'
@end example

@findex CONCAT_WS()
@item CONCAT_WS(separator, str1, str2,...)

@code{CONCAT_WS()} stands for CONCAT With Separator and is a special form of
@code{CONCAT()}.  The first argument is the separator for the rest of the
arguments. The separator can be a string as well as the rest of the
arguments. If the separator is @code{NULL}, the result will be @code{NULL}.
The function will skip any @code{NULL}s and empty strings, after the
separator argument. The separator will be added between the strings to be
concatenated:

@example
mysql> SELECT CONCAT_WS(",","First name","Second name","Last Name");
       -> 'First name,Second name,Last Name'
mysql> SELECT CONCAT_WS(",","First name",NULL,"Last Name");
       -> 'First name,Last Name'
@end example

@findex LENGTH()
@findex OCTET_LENGTH()
@findex CHAR_LENGTH()
@findex CHARACTER_LENGTH()
@item LENGTH(str)
@itemx OCTET_LENGTH(str)
@itemx CHAR_LENGTH(str)
@itemx CHARACTER_LENGTH(str)
Returns the length of the string @code{str}:

@example
mysql> SELECT LENGTH('text');
        -> 4
mysql> SELECT OCTET_LENGTH('text');
        -> 4
@end example

Note that for @code{CHAR_LENGTH()} and @code{CHARACTER_LENGTH()}, multi-byte
characters are only counted once.

@findex BIT_LENGTH()
@item BIT_LENGTH(str)
Returns the length of the string @code{str} in bits:

@example
mysql> SELECT BIT_LENGTH('text');
        -> 32
@end example

@findex LOCATE()
@findex POSITION()
@item LOCATE(substr,str)
@itemx POSITION(substr IN str)
Returns the position of the first occurrence of substring @code{substr}
in string @code{str}. Returns @code{0} if @code{substr} is not in @code{str}:

@example
mysql> SELECT LOCATE('bar', 'foobarbar');
        -> 4
mysql> SELECT LOCATE('xbar', 'foobar');
        -> 0
@end example

This function is multi-byte safe.  In MySQL 3.23 this function is case
insensitive, while in 4.0 it's only case-insensitive if either argument is
a binary string.

@findex LOCATE()
@item LOCATE(substr,str,pos)
Returns the position of the first occurrence of substring @code{substr} in
string @code{str}, starting at position @code{pos}.
Returns @code{0} if @code{substr} is not in @code{str}:

@example
mysql> SELECT LOCATE('bar', 'foobarbar',5);
        -> 7
@end example

This function is multi-byte safe.  In MySQL 3.23 this function is case
insensitive, while in 4.0 it's only case-insensitive if either argument is
a binary string.

@findex INSTR()
@item INSTR(str,substr)
Returns the position of the first occurrence of substring @code{substr} in
string @code{str}. This is the same as the two-argument form of
@code{LOCATE()}, except that the arguments are swapped:

@example
mysql> SELECT INSTR('foobarbar', 'bar');
        -> 4
mysql> SELECT INSTR('xbar', 'foobar');
        -> 0
@end example

This function is multi-byte safe.  In MySQL 3.23 this function is case
insensitive, while in 4.0 it's only case-insensitive if either argument is
a binary string.

@findex LPAD()
@item LPAD(str,len,padstr)
Returns the string @code{str}, left-padded with the string @code{padstr}
until @code{str} is @code{len} characters long. If @code{str} is longer
than @code{len'} then it will be shortened to @code{len} characters.

@example
mysql> SELECT LPAD('hi',4,'??');
        -> '??hi'
@end example

@findex RPAD()
@item RPAD(str,len,padstr)
Returns the string @code{str}, right-padded with the string
@code{padstr} until @code{str} is @code{len} characters long.  If
@code{str} is longer than @code{len'} then it will be shortened to
@code{len} characters.

@example
mysql> SELECT RPAD('hi',5,'?');
        -> 'hi???'
@end example

@findex LEFT()
@item LEFT(str,len)
Returns the leftmost @code{len} characters from the string @code{str}:

@example
mysql> SELECT LEFT('foobarbar', 5);
        -> 'fooba'
@end example

This function is multi-byte safe.

@findex RIGHT()
@item RIGHT(str,len)
Returns the rightmost @code{len} characters from the string @code{str}:

@example
mysql> SELECT RIGHT('foobarbar', 4);
        -> 'rbar'
@end example

This function is multi-byte safe.

@findex SUBSTRING()
@findex MID()
@item SUBSTRING(str,pos,len)
@itemx SUBSTRING(str FROM pos FOR len)
@itemx MID(str,pos,len)
Returns a substring @code{len} characters long from string @code{str},
starting at position @code{pos}.
The variant form that uses @code{FROM} is ANSI SQL92 syntax:

@example
mysql> SELECT SUBSTRING('Quadratically',5,6);
        -> 'ratica'
@end example

This function is multi-byte safe.

@findex SUBSTRING()
@item SUBSTRING(str,pos)
@item SUBSTRING(str FROM pos)
Returns a substring from string @code{str} starting at position @code{pos}:

@example
mysql> SELECT SUBSTRING('Quadratically',5);
        -> 'ratically'
mysql> SELECT SUBSTRING('foobarbar' FROM 4);
        -> 'barbar'
@end example

This function is multi-byte safe.

@findex SUBSTRING_INDEX()
@item SUBSTRING_INDEX(str,delim,count)
Returns the substring from string @code{str} before @code{count}
occurrences of the delimiter @code{delim}.
If @code{count} is positive, everything to the left of the final delimiter
(counting from the left) is returned.
If @code{count} is negative, everything to the right of the final delimiter
(counting from the right) is returned:

@example
mysql> SELECT SUBSTRING_INDEX('www.mysql.com', '.', 2);
        -> 'www.mysql'
mysql> SELECT SUBSTRING_INDEX('www.mysql.com', '.', -2);
        -> 'mysql.com'
@end example

This function is multi-byte safe.

@findex LTRIM()
@item LTRIM(str)
Returns the string @code{str} with leading space characters removed:

@example
mysql> SELECT LTRIM('  barbar');
        -> 'barbar'
@end example

@findex RTRIM()
@item RTRIM(str)
Returns the string @code{str} with trailing space characters removed:

@example
mysql> SELECT RTRIM('barbar   ');
        -> 'barbar'
@end example

This function is multi-byte safe.

@findex TRIM()
@item TRIM([[BOTH | LEADING | TRAILING] [remstr] FROM] str)
Returns the string @code{str} with all @code{remstr} prefixes and/or suffixes
removed. If none of the specifiers @code{BOTH}, @code{LEADING} or
@code{TRAILING} are given, @code{BOTH} is assumed. If @code{remstr} is not
specified, spaces are removed:

@example
mysql> SELECT TRIM('  bar   ');
        -> 'bar'
mysql> SELECT TRIM(LEADING 'x' FROM 'xxxbarxxx');
        -> 'barxxx'
mysql> SELECT TRIM(BOTH 'x' FROM 'xxxbarxxx');
        -> 'bar'
mysql> SELECT TRIM(TRAILING 'xyz' FROM 'barxxyz');
        -> 'barx'
@end example

This function is multi-byte safe.

@findex SOUNDEX()
@item SOUNDEX(str)
Returns a soundex string from @code{str}. Two strings that sound almost the
same should have identical soundex strings. A standard soundex string
is 4 characters long, but the @code{SOUNDEX()} function returns an
arbitrarily long string. You can use @code{SUBSTRING()} on the result to get
a standard soundex string.  All non-alphanumeric characters are ignored
in the given string. All international alpha characters outside the A-Z range
are treated as vowels:

@example
mysql> SELECT SOUNDEX('Hello');
        -> 'H400'
mysql> SELECT SOUNDEX('Quadratically');
        -> 'Q36324'
@end example

@findex SPACE()
@item SPACE(N)
Returns a string consisting of @code{N} space characters:

@example
mysql> SELECT SPACE(6);
        -> '      '
@end example

@findex REPLACE()
@item REPLACE(str,from_str,to_str)
Returns the string @code{str} with all all occurrences of the string
@code{from_str} replaced by the string @code{to_str}:

@example
mysql> SELECT REPLACE('www.mysql.com', 'w', 'Ww');
        -> 'WwWwWw.mysql.com'
@end example

This function is multi-byte safe.

@findex REPEAT()
@item REPEAT(str,count)
Returns a string consisting of the string @code{str} repeated @code{count}
times. If @code{count <= 0}, returns an empty string. Returns @code{NULL} if
@code{str} or @code{count} are @code{NULL}:

@example
mysql> SELECT REPEAT('MySQL', 3);
        -> 'MySQLMySQLMySQL'
@end example

@findex REVERSE()
@item REVERSE(str)
Returns the string @code{str} with the order of the characters reversed:

@example
mysql> SELECT REVERSE('abc');
        -> 'cba'
@end example

This function is multi-byte safe.

@findex INSERT()
@item INSERT(str,pos,len,newstr)
Returns the string @code{str}, with the substring beginning at position
@code{pos} and @code{len} characters long replaced by the string
@code{newstr}:

@example
mysql> SELECT INSERT('Quadratic', 3, 4, 'What');
        -> 'QuWhattic'
@end example

This function is multi-byte safe.

@findex ELT()
@item ELT(N,str1,str2,str3,...)
Returns @code{str1} if @code{N} = @code{1}, @code{str2} if @code{N} =
@code{2}, and so on.  Returns @code{NULL} if @code{N} is less than @code{1}
or greater than the number of arguments.  @code{ELT()} is the complement of
@code{FIELD()}:

@example
mysql> SELECT ELT(1, 'ej', 'Heja', 'hej', 'foo');
        -> 'ej'
mysql> SELECT ELT(4, 'ej', 'Heja', 'hej', 'foo');
        -> 'foo'
@end example

@findex FIELD()
@item FIELD(str,str1,str2,str3,...)
Returns the index of @code{str} in the @code{str1}, @code{str2},
@code{str3}, @code{...} list.
Returns @code{0} if @code{str} is not found.
@code{FIELD()} is the complement of @code{ELT()}:

@example
mysql> SELECT FIELD('ej', 'Hej', 'ej', 'Heja', 'hej', 'foo');
        -> 2
mysql> SELECT FIELD('fo', 'Hej', 'ej', 'Heja', 'hej', 'foo');
        -> 0
@end example

@findex FIND_IN_SET()
@item FIND_IN_SET(str,strlist)
Returns a value @code{1} to @code{N} if the string @code{str} is in the list
@code{strlist} consisting of @code{N} substrings. A string list is a string
composed of substrings separated by @samp{,} characters. If the first
argument is a constant string and the second is a column of type @code{SET},
the @code{FIND_IN_SET()} function is optimised to use bit arithmetic!
Returns @code{0} if @code{str} is not in @code{strlist} or if @code{strlist}
is the empty string.  Returns @code{NULL} if either argument is @code{NULL}.
This function will not work properly if the first argument contains a
@samp{,}:

@example
mysql> SELECT FIND_IN_SET('b','a,b,c,d');
        -> 2
@end example

@findex MAKE_SET()
@item MAKE_SET(bits,str1,str2,...)
Returns a set (a string containing substrings separated by @samp{,}
characters) consisting of the strings that have the corresponding bit in
@code{bits} set.  @code{str1} corresponds to bit 0, @code{str2} to bit 1,
etc.  @code{NULL} strings in @code{str1}, @code{str2}, @code{...}
are not appended to the result:

@example
mysql> SELECT MAKE_SET(1,'a','b','c');
        -> 'a'
mysql> SELECT MAKE_SET(1 | 4,'hello','nice','world');
        -> 'hello,world'
mysql> SELECT MAKE_SET(0,'a','b','c');
        -> ''
@end example

@findex EXPORT_SET()
@item EXPORT_SET(bits,on,off,[separator,[number_of_bits]])
Returns a string where for every bit set in 'bit', you get an 'on' string
and for every reset bit you get an 'off' string. Each string is separated
with 'separator' (default ',') and only 'number_of_bits' (default 64) of
'bits' is used:

@example
mysql> SELECT EXPORT_SET(5,'Y','N',',',4)
        -> Y,N,Y,N
@end example

@findex LCASE()
@findex LOWER()
@item LCASE(str)
@itemx LOWER(str)
Returns the string @code{str} with all characters changed to lowercase
according to the current character set mapping (the default is ISO-8859-1
Latin1):

@example
mysql> SELECT LCASE('QUADRATICALLY');
        -> 'quadratically'
@end example

This function is multi-byte safe.

@findex UCASE()
@findex UPPER()
@item UCASE(str)
@itemx UPPER(str)
Returns the string @code{str} with all characters changed to uppercase
according to the current character set mapping (the default is ISO-8859-1
Latin1):

@example
mysql> SELECT UCASE('Hej');
        -> 'HEJ'
@end example

This function is multi-byte safe.

@findex FILE
@findex LOAD_FILE()
@item LOAD_FILE(file_name)
Reads the file and returns the file contents as a string.  The file
must be on the server, you must specify the full pathname to the
file, and you must have the @strong{file} privilege.  The file must
be readable by all and be smaller than @code{max_allowed_packet}.

If the file doesn't exist or can't be read due to one of the above reasons,
the function returns @code{NULL}:

@example
mysql> UPDATE tbl_name
           SET blob_column=LOAD_FILE("/tmp/picture")
           WHERE id=1;
@end example
@end table

If you are not using MySQL Version 3.23, you have to do the reading
of the file inside your application and create an @code{INSERT} statement
to update the database with the file information. One way to do this, if
you are using the MySQL++ library, can be found at
@uref{http://www.mysql.com/documentation/mysql++/mysql++-examples.html}.

MySQL automatically converts numbers to strings as necessary, and
vice-versa:

@example
mysql> SELECT 1+"1";
        -> 2
mysql> SELECT CONCAT(2,' test');
        -> '2 test'
@end example

If you want to convert a number to a string explicitly, pass it as the
argument to @code{CONCAT()}.

If a string function is given a binary string as an argument, the resulting
string is also a binary string.  A number converted to a string is treated as
a binary string.  This only affects comparisons.


@menu
* String comparison functions::  String Comparison Functions
* Case Sensitivity Operators::  Case Sensitivity
@end menu

@node String comparison functions, Case Sensitivity Operators, String functions, String functions
@subsubsection String Comparison Functions

@findex string comparison functions
@findex functions, string comparison

@cindex case sensitivity, in string comparisons
@cindex string comparisons, case sensitivity

Normally, if any expression in a string comparison is case-sensitive, the
comparison is performed in case-sensitive fashion.

@table @code
@findex LIKE
@item expr LIKE pat [ESCAPE 'escape-char']
Pattern matching using
SQL simple regular expression comparison. Returns @code{1} (TRUE) or @code{0}
(FALSE).  With @code{LIKE} you can use the following two wildcard characters
in the pattern:

@multitable @columnfractions .10 .60
@item @strong{Char} @tab @strong{Description}
@item @code{%} @tab Matches any number of characters, even zero characters
@item @code{_} @tab Matches exactly one character
@end multitable

@example
mysql> SELECT 'David!' LIKE 'David_';
        -> 1
mysql> SELECT 'David!' LIKE '%D%v%';
        -> 1
mysql> select TRUNCATE(-1,999,1);
        -> -1.9
@end example

Starting from MySQL 3.23.51 all numbers are rounded towards zero.

If @code{D} is negative, then the whole part of the number is zeroed out:

@example
mysql> select truncate(122,-2);
       -> 100
@end example

To test for literal instances of a wildcard character, precede the character
with the escape character.  If you don't specify the @code{ESCAPE} character,
@samp{\} is assumed:

@multitable @columnfractions .10 .60
@item @strong{String} @tab @strong{Description}
@item @code{\%} @tab Matches one @code{%} character
@item @code{\_} @tab Matches one @code{_} character
@end multitable

@example
mysql> SELECT 'David!' LIKE 'David\_';
        -> 0
mysql> SELECT 'David_' LIKE 'David\_';
        -> 1
@end example

To specify a different escape character, use the @code{ESCAPE} clause:

@example
mysql> SELECT 'David_' LIKE 'David|_' ESCAPE '|';
        -> 1
@end example

The following two statements illustrate that string comparisons are
case-insensitive unless one of the operands is a binary string:

@example
mysql> SELECT 'abc' LIKE 'ABC';
        -> 1
mysql> SELECT 'abc' LIKE BINARY 'ABC';
        -> 0
@end example

@code{LIKE} is allowed on numeric expressions! (This is a MySQL
extension to the ANSI SQL @code{LIKE}.)

@example
mysql> SELECT 10 LIKE '1%';
        -> 1
@end example

Note: Because MySQL uses the C escape syntax in strings (for example,
@samp{\n}), you must double any @samp{\} that you use in your @code{LIKE}
strings.  For example, to search for @samp{\n}, specify it as @samp{\\n}.  To
search for @samp{\}, specify it as @samp{\\\\} (the backslashes are stripped
once by the parser and another time when the pattern match is done, leaving
a single backslash to be matched).

@findex NOT LIKE
@item expr NOT LIKE pat [ESCAPE 'escape-char']
Same as @code{NOT (expr LIKE pat [ESCAPE 'escape-char'])}.

@cindex mSQL compatibility
@cindex compatibility, with mSQL
@findex REGEXP
@findex RLIKE
@item expr REGEXP pat
@itemx expr RLIKE pat
Performs a pattern match of a string expression @code{expr} against a pattern
@code{pat}.  The pattern can be an extended regular expression.
@xref{Regexp}.  Returns @code{1} if @code{expr} matches @code{pat}, otherwise
returns @code{0}.  @code{RLIKE} is a synonym for @code{REGEXP}, provided for
@code{mSQL} compatibility. Note: Because MySQL uses the C escape
syntax in strings (for example, @samp{\n}), you must double any @samp{\} that
you use in your @code{REGEXP} strings.  As of MySQL Version 3.23.4,
@code{REGEXP} is case-insensitive for normal (not binary) strings:

@example
mysql> SELECT 'Monty!' REGEXP 'm%y%%';
        -> 0
mysql> SELECT 'Monty!' REGEXP '.*';
        -> 1
mysql> SELECT 'new*\n*line' REGEXP 'new\\*.\\*line';
        -> 1
mysql> SELECT "a" REGEXP "A", "a" REGEXP BINARY "A";
        -> 1  0
mysql> SELECT "a" REGEXP "^[a-d]";
        -> 1
@end example

@item
@code{REGEXP} and @code{RLIKE} use the current character set (ISO-8859-1
Latin1 by default) when deciding the type of a character.

@findex NOT REGEXP
@item expr NOT REGEXP pat
@itemx expr NOT RLIKE pat
Same as @code{NOT (expr REGEXP pat)}.

@findex STRCMP()
@item STRCMP(expr1,expr2)
@code{STRCMP()}
returns @code{0} if the strings are the same, @code{-1} if the first
argument is smaller than the second according to the current sort order,
and @code{1} otherwise:

@example
mysql> SELECT STRCMP('text', 'text2');
        -> -1
mysql> SELECT STRCMP('text2', 'text');
        -> 1
mysql> SELECT STRCMP('text', 'text');
        -> 0
@end example

@findex MATCH ... AGAINST()
@item MATCH (col1,col2,...) AGAINST (expr)
@itemx MATCH (col1,col2,...) AGAINST (expr IN BOOLEAN MODE)
@code{MATCH ... AGAINST()} is used for full-text search and returns
relevance - similarity measure between the text in columns
@code{(col1,col2,...)} and the query @code{expr}. Relevance is a
positive floating-point number. Zero relevance means no similarity.
@code{MATCH ... AGAINST()} is available in MySQL version
3.23.23 or later. @code{IN BOOLEAN MODE} extension was added in version
4.0.1. For details and usage examples, see @ref{Fulltext Search}.
@end table

@node Case Sensitivity Operators,  , String comparison functions, String functions
@subsubsection Case Sensitivity

@findex casts

@cindex cast operators
@cindex operators, cast

@table @code
@findex BINARY
@item @code{BINARY}
The @code{BINARY} operator casts the string following it to a binary string.
This is an easy way to force a column comparison to be case-sensitive even
if the column isn't defined as @code{BINARY} or @code{BLOB}:
@example
mysql> SELECT "a" = "A";
        -> 1
mysql> SELECT BINARY "a" = "A";
        -> 0
@end example

@code{BINARY string} is a shorthand for @code{CAST(string AS BINARY)}.
@xref{Cast Functions}.
@code{BINARY} was introduced in MySQL Version 3.23.0.

Note that in some context MySQL will not be able to use the
index efficiently when you cast an indexed column to @code{BINARY}.
@end table

If you want to compare a blob case-insensitively you can always convert
the blob to upper case before doing the comparison:

@example
SELECT 'A' LIKE UPPER(blob_col) FROM table_name;
@end example

We plan to soon introduce casting between different character sets to
make string comparison even more flexible.


@node Numeric Functions, Date and time functions, String functions, Functions
@subsection Numeric Functions

@menu
* Arithmetic functions::        Arithmetic Operations
* Mathematical functions::      Mathematical Functions
@end menu


@node Arithmetic functions, Mathematical functions, Numeric Functions, Numeric Functions
@subsubsection Arithmetic Operations

@cindex operators, cast
The usual arithmetic operators are available. Note that in the case of
@samp{-}, @samp{+}, and @samp{*}, the result is calculated with
@code{BIGINT} (64-bit) precision if both arguments are integers!
If one of the argument is an unsigned integer, and the other argument
is also an integer, the result will be an unsigned integer.
@xref{Cast Functions}.

@cindex operations, arithmetic
@cindex arithmetic expressions
@table @code
@findex + (addition)
@findex addition (+)
@item +
Addition:
@example
mysql> SELECT 3+5;
        -> 8
@end example

@findex - (subtraction)
@findex subtraction (-)
@item -
Subtraction:
@example
mysql> SELECT 3-5;
        -> -2
@end example

@findex * (multiplication)
@findex multiplication (*)
@item *
Multiplication:
@example
mysql> SELECT 3*5;
        -> 15
mysql> SELECT 18014398509481984*18014398509481984.0;
        -> 324518553658426726783156020576256.0
mysql> SELECT 18014398509481984*18014398509481984;
        -> 0
@end example

The result of the last expression is incorrect because the result of the
integer multiplication exceeds the 64-bit range of @code{BIGINT}
calculations.

@findex / (division)
@findex division (/)
@item /
Division:
@example
mysql> SELECT 3/5;
        -> 0.60
@end example

Division by zero produces a @code{NULL} result:

@example
mysql> SELECT 102/(1-1);
        -> NULL
@end example

A division will be calculated with @code{BIGINT} arithmetic only if performed
in a context where its result is converted to an integer!
@end table


@node Mathematical functions,  , Arithmetic functions, Numeric Functions
@subsubsection Mathematical Functions
All mathematical functions return @code{NULL} in case of an error.

@findex mathematical functions
@findex functions, mathematical

@table @code
@findex - (unary minus)
@findex minus, unary (-)
@findex unary minus (-)
@item -
Unary minus. Changes the sign of the argument:
@example
mysql> SELECT - 2;
        -> -2
@end example

Note that if this operator is used with a @code{BIGINT}, the return value is a
@code{BIGINT}!  This means that you should avoid using @code{-} on integers that
may have the value of @code{-2^63}!

@findex ABS()
@item ABS(X)
Returns the absolute value of @code{X}:
@example
mysql> SELECT ABS(2);
        -> 2
mysql> SELECT ABS(-32);
        -> 32
@end example

This function is safe to use with @code{BIGINT} values.

@findex SIGN()
@item SIGN(X)
Returns the sign of the argument as @code{-1}, @code{0}, or @code{1}, depending
on whether @code{X} is negative, zero, or positive:
@example
mysql> SELECT SIGN(-32);
        -> -1
mysql> SELECT SIGN(0);
        -> 0
mysql> SELECT SIGN(234);
        -> 1
@end example

@findex MOD()
@findex % (modulo)
@findex modulo (%)
@item MOD(N,M)
@itemx %
Modulo (like the @code{%} operator in C).
Returns the remainder of @code{N} divided by @code{M}:
@example
mysql> SELECT MOD(234, 10);
        -> 4
mysql> SELECT 253 % 7;
        -> 1
mysql> SELECT MOD(29,9);
        -> 2
@end example

This function is safe to use with @code{BIGINT} values.

@findex FLOOR()
@item FLOOR(X)
Returns the largest integer value not greater than @code{X}:
@example
mysql> SELECT FLOOR(1.23);
        -> 1
mysql> SELECT FLOOR(-1.23);
        -> -2
@end example

Note that the return value is converted to a @code{BIGINT}!

@findex CEILING()
@item CEILING(X)
Returns the smallest integer value not less than @code{X}:
@example
mysql> SELECT CEILING(1.23);
        -> 2
mysql> SELECT CEILING(-1.23);
        -> -1
@end example

Note that the return value is converted to a @code{BIGINT}!

@findex ROUND()
@item ROUND(X)
Returns the argument @code{X}, rounded to the nearest integer:
@example
mysql> SELECT ROUND(-1.23);
        -> -1
mysql> SELECT ROUND(-1.58);
        -> -2
mysql> SELECT ROUND(1.58);
        -> 2
@end example

Note that the behavior of @code{ROUND()} when the argument
is half way between two integers depends on the C library
implementation.  Some round to the nearest even number,
always up, always down, or always toward zero.  If you need
one kind of rounding, you should use a well-defined function
like @code{TRUNCATE()} or @code{FLOOR()} instead.

@findex ROUND()
@item ROUND(X,D)
Returns the argument @code{X}, rounded to a number with @code{D} decimals.
If @code{D} is @code{0}, the result will have no decimal point or fractional
part:

@example
mysql> SELECT ROUND(1.298, 1);
        -> 1.3
mysql> SELECT ROUND(1.298, 0);
        -> 1
@end example

@findex EXP()
@item EXP(X)
Returns the value of @code{e} (the base of natural logarithms) raised to
the power of @code{X}:
@example
mysql> SELECT EXP(2);
        -> 7.389056
mysql> SELECT EXP(-2);
        -> 0.135335
@end example
@findex LOG()
@item LOG(X)
Returns the natural logarithm of @code{X}:
@example
mysql> SELECT LOG(2);
        -> 0.693147
mysql> SELECT LOG(-2);
        -> NULL
@end example
If you want the log of a number @code{X} to some arbitary base @code{B}, use
the formula @code{LOG(X)/LOG(B)}.

@findex LOG10()
@item LOG10(X)
Returns the base-10 logarithm of @code{X}:
@example
mysql> SELECT LOG10(2);
        -> 0.301030
mysql> SELECT LOG10(100);
        -> 2.000000
mysql> SELECT LOG10(-100);
        -> NULL
@end example

@findex POW()
@findex POWER()
@item POW(X,Y)
@itemx POWER(X,Y)
Returns the value of @code{X} raised to the power of @code{Y}:
@example
mysql> SELECT POW(2,2);
        -> 4.000000
mysql> SELECT POW(2,-2);
        -> 0.250000
@end example

@findex SQRT()
@item SQRT(X)
Returns the non-negative square root of @code{X}:
@example
mysql> SELECT SQRT(4);
        -> 2.000000
mysql> SELECT SQRT(20);
        -> 4.472136
@end example

@findex PI()
@item PI()
Returns the value of PI. The default shown number of decimals is 5, but
MySQL internally uses the full double precession for PI.
@example
mysql> SELECT PI();
        -> 3.141593
mysql> SELECT PI()+0.000000000000000000;
        -> 3.141592653589793116
@end example

@findex COS()
@item COS(X)
Returns the cosine of @code{X}, where @code{X} is given in radians:
@example
mysql> SELECT COS(PI());
        -> -1.000000
@end example

@findex SIN()
@item SIN(X)
Returns the sine of @code{X}, where @code{X} is given in radians:
@example
mysql> SELECT SIN(PI());
        -> 0.000000
@end example

@findex TAN()
@item TAN(X)
Returns the tangent of @code{X}, where @code{X} is given in radians:
@example
mysql> SELECT TAN(PI()+1);
        -> 1.557408
@end example

@findex ACOS()
@item ACOS(X)
Returns the arc cosine of @code{X}, that is, the value whose cosine is
@code{X}. Returns @code{NULL} if @code{X} is not in the range @code{-1} to
@code{1}:
@example
mysql> SELECT ACOS(1);
        -> 0.000000
mysql> SELECT ACOS(1.0001);
        -> NULL
mysql> SELECT ACOS(0);
        -> 1.570796
@end example

@findex ASIN()
@item ASIN(X)
Returns the arc sine of @code{X}, that is, the value whose sine is
@code{X}. Returns @code{NULL} if @code{X} is not in the range @code{-1} to
@code{1}:
@example
mysql> SELECT ASIN(0.2);
        -> 0.201358
mysql> SELECT ASIN('foo');
        -> 0.000000
@end example

@findex ATAN()
@item ATAN(X)
Returns the arc tangent of @code{X}, that is, the value whose tangent is
@code{X}:
@example
mysql> SELECT ATAN(2);
        -> 1.107149
mysql> SELECT ATAN(-2);
        -> -1.107149
@end example

@findex ATAN2()
@item ATAN(Y,X)
@itemx ATAN2(Y,X)
Returns the arc tangent of the two variables @code{X} and @code{Y}. It is
similar to calculating the arc tangent of @code{Y / X}, except that the
signs of both arguments are used to determine the quadrant of the
result:
@example
mysql> SELECT ATAN(-2,2);
        -> -0.785398
mysql> SELECT ATAN2(PI(),0);
        -> 1.570796
@end example

@findex COT()
@item COT(X)
Returns the cotangent of @code{X}:
@example
mysql> SELECT COT(12);
        -> -1.57267341
mysql> SELECT COT(0);
        -> NULL
@end example

@findex RAND()
@item RAND()
@itemx RAND(N)
Returns a random floating-point value in the range @code{0} to @code{1.0}.
If an integer argument @code{N} is specified, it is used as the seed value:
@example
mysql> SELECT RAND();
        -> 0.9233482386203
mysql> SELECT RAND(20);
        -> 0.15888261251047
mysql> SELECT RAND(20);
        -> 0.15888261251047
mysql> SELECT RAND();
        -> 0.63553050033332
mysql> SELECT RAND();
        -> 0.70100469486881
@end example

You can't use a column with @code{RAND()} values in an @code{ORDER BY}
clause, because @code{ORDER BY} would evaluate the column multiple times.
In MySQL Version 3.23, you can, however, do:
@code{SELECT * FROM table_name ORDER BY RAND()}

This is useful to get a random sample of a set @code{SELECT * FROM
table1,table2 WHERE a=b AND c<d ORDER BY RAND() LIMIT 1000}.

Note that a @code{RAND()} in a @code{WHERE} clause will be re-evaluated
every time the @code{WHERE} is executed.

@code{RAND()} is not meant to be a perfect random generator, but instead a
fast way to generate ad hoc random numbers that will be portable between
platforms for the same MySQL version.

@findex LEAST()
@item LEAST(X,Y,...)
With two or more arguments, returns the smallest (minimum-valued) argument.
The arguments are compared using the following rules:

@itemize @bullet
@item
If the return value is used in an @code{INTEGER} context, or all arguments
are integer-valued, they are compared as integers.

@item
If the return value is used in a @code{REAL} context, or all arguments are
real-valued, they are compared as reals.

@item
If any argument is a case-sensitive string, the arguments are compared
as case-sensitive strings.

@item
In other cases, the arguments are compared as case-insensitive strings:
@end itemize

@example
mysql> SELECT LEAST(2,0);
        -> 0
mysql> SELECT LEAST(34.0,3.0,5.0,767.0);
        -> 3.0
mysql> SELECT LEAST("B","A","C");
        -> "A"
@end example
In MySQL versions prior to Version 3.22.5, you can use @code{MIN()}
instead of @code{LEAST}.

@findex GREATEST()
@item GREATEST(X,Y,...)
Returns the largest (maximum-valued) argument.
The arguments are compared using the same rules as for @code{LEAST}:
@example
mysql> SELECT GREATEST(2,0);
        -> 2
mysql> SELECT GREATEST(34.0,3.0,5.0,767.0);
        -> 767.0
mysql> SELECT GREATEST("B","A","C");
        -> "C"
@end example
In MySQL versions prior to Version 3.22.5, you can use @code{MAX()}
instead of @code{GREATEST}.

@findex DEGREES()
@item DEGREES(X)
Returns the argument @code{X}, converted from radians to degrees:
@example
mysql> SELECT DEGREES(PI());
        -> 180.000000
@end example

@findex RADIANS()
@item RADIANS(X)
Returns the argument @code{X}, converted from degrees to radians:
@example
mysql> SELECT RADIANS(90);
        -> 1.570796
@end example

@findex TRUNCATE()
@item TRUNCATE(X,D)
Returns the number @code{X}, truncated to @code{D} decimals.  If @code{D}
is @code{0}, the result will have no decimal point or fractional part:
@example
mysql> SELECT TRUNCATE(1.223,1);
        -> 1.2
mysql> SELECT TRUNCATE(1.999,1);
        -> 1.9
mysql> SELECT TRUNCATE(1.999,0);
        -> 1
@end example

Note that as decimal numbers are normally not stored as exact numbers in
computers, but as double values, you may be fooled by the following
result:

@cindex rounding errors
@example
mysql> SELECT TRUNCATE(10.28*100,0);
       -> 1027
@end example

The above happens because 10.28 is actually stored as something like
10.2799999999999999.
@end table

@node Date and time functions, Cast Functions, Numeric Functions, Functions
@subsection Date and Time Functions

@findex date and time functions
@findex functions, date and time

See @ref{Date and time types} for a description of the range of values
each type has and the valid formats in which date and time values may be
specified.

Here is an example that uses date functions.  The following query selects
all records with a @code{date_col} value from within the last 30 days:

@example
mysql> SELECT something FROM tbl_name
           WHERE TO_DAYS(NOW()) - TO_DAYS(date_col) <= 30;
@end example

@table @code
@findex DAYOFWEEK()
@item DAYOFWEEK(date)
Returns the weekday index

for @code{date} (@code{1} = Sunday, @code{2} = Monday, ... @code{7} =
Saturday).  These index values correspond to the ODBC standard:

@example
mysql> SELECT DAYOFWEEK('1998-02-03');
        -> 3
@end example

@findex WEEKDAY()
@item WEEKDAY(date)
Returns the weekday index for
@code{date} (@code{0} = Monday, @code{1} = Tuesday, ... @code{6} = Sunday):

@example
mysql> SELECT WEEKDAY('1997-10-04 22:23:00');
        -> 5
mysql> SELECT WEEKDAY('1997-11-05');
        -> 2
@end example

@findex DAYOFMONTH()
@item DAYOFMONTH(date)
Returns the day of the month for @code{date}, in the range @code{1} to
@code{31}:

@example
mysql> SELECT DAYOFMONTH('1998-02-03');
        -> 3
@end example

@findex DAYOFYEAR()
@item DAYOFYEAR(date)
Returns the day of the year for @code{date}, in the range @code{1} to
@code{366}:

@example
mysql> SELECT DAYOFYEAR('1998-02-03');
        -> 34
@end example

@findex MONTH()
@item MONTH(date)
Returns the month for @code{date}, in the range @code{1} to @code{12}:

@example
mysql> SELECT MONTH('1998-02-03');
        -> 2
@end example

@findex DAYNAME()
@item DAYNAME(date)
Returns the name of the weekday for @code{date}:

@example
mysql> SELECT DAYNAME("1998-02-05");
        -> 'Thursday'
@end example

@findex MONTHNAME()
@item MONTHNAME(date)
Returns the name of the month for @code{date}:

@example
mysql> SELECT MONTHNAME("1998-02-05");
        -> 'February'
@end example

@findex QUARTER()
@item QUARTER(date)
Returns the quarter of the year for @code{date}, in the range @code{1}
to @code{4}:

@example
mysql> SELECT QUARTER('98-04-01');
        -> 2
@end example

@findex WEEK()
@item WEEK(date)
@itemx WEEK(date,first)
With a single argument, returns the week for @code{date}, in the range
@code{0} to @code{53} (yes, there may be the beginnings of a week 53),
for locations where Sunday is the first day of the week.  The
two-argument form of @code{WEEK()} allows you to specify whether the
week starts on Sunday or Monday.  The week starts on Sunday if the
second argument is @code{0}, on Monday if the second argument is
@code{1}:

@example
mysql> SELECT WEEK('1998-02-20');
        -> 7
mysql> SELECT WEEK('1998-02-20',0);
        -> 7
mysql> SELECT WEEK('1998-02-20',1);
        -> 8
mysql> SELECT WEEK('1998-12-31',1);
        -> 53
@end example

Note: in Version 4.0, @code{WEEK(#,0)} was changed to match the
calendar in the USA.

@findex YEAR()
@item YEAR(date)
Returns the year for @code{date}, in the range @code{1000} to @code{9999}:

@example
mysql> SELECT YEAR('98-02-03');
        -> 1998
@end example

@item YEARWEEK(date)
@itemx YEARWEEK(date,first)
Returns year and week for a date.  The second arguments works exactly
like the second argument to @code{WEEK()}.  Note that the year may be
different from the year in the date argument for the first and the last
week of the year:

@example
mysql> SELECT YEARWEEK('1987-01-01');
        -> 198653
@end example

@findex HOUR()
@item HOUR(time)
Returns the hour for @code{time}, in the range @code{0} to @code{23}:

@example
mysql> SELECT HOUR('10:05:03');
        -> 10
@end example

@findex MINUTE()
@item MINUTE(time)
Returns the minute for @code{time}, in the range @code{0} to @code{59}:

@example
mysql> SELECT MINUTE('98-02-03 10:05:03');
        -> 5
@end example

@findex SECOND()
@item SECOND(time)
Returns the second for @code{time}, in the range @code{0} to @code{59}:

@example
mysql> SELECT SECOND('10:05:03');
        -> 3
@end example

@findex PERIOD_ADD()
@item PERIOD_ADD(P,N)
Adds @code{N} months to period @code{P} (in the format @code{YYMM} or
@code{YYYYMM}). Returns a value in the format @code{YYYYMM}.

Note that the period argument @code{P} is @strong{not} a date value:

@example
mysql> SELECT PERIOD_ADD(9801,2);
        -> 199803
@end example

@findex PERIOD_DIFF()
@item PERIOD_DIFF(P1,P2)
Returns the number of months between periods @code{P1} and @code{P2}.
@code{P1} and @code{P2} should be in the format @code{YYMM} or @code{YYYYMM}.

Note that the period arguments @code{P1} and @code{P2} are @strong{not}
date values:

@example
mysql> SELECT PERIOD_DIFF(9802,199703);
        -> 11
@end example

@findex DATE_ADD()
@findex DATE_SUB()
@findex ADDDATE()
@findex SUBDATE()
@findex EXTRACT()
@item DATE_ADD(date,INTERVAL expr type)
@itemx DATE_SUB(date,INTERVAL expr type)
@itemx ADDDATE(date,INTERVAL expr type)
@itemx SUBDATE(date,INTERVAL expr type)

These functions perform date arithmetic.  They are new for MySQL
Version 3.22.  @code{ADDDATE()} and @code{SUBDATE()} are synonyms for
@code{DATE_ADD()} and @code{DATE_SUB()}.

In MySQL Version 3.23, you can use @code{+} and @code{-} instead of
@code{DATE_ADD()} and @code{DATE_SUB()} if the expression on the right side is
a date or datetime column. (See example below.)

@code{date} is a @code{DATETIME} or @code{DATE} value specifying the starting
date.  @code{expr} is an expression specifying the interval value to be added
or subtracted from the starting date.  @code{expr} is a string; it may start
with a @samp{-} for negative intervals.  @code{type} is a keyword indicating
how the expression should be interpreted.

The related function @code{EXTRACT(type FROM date)} returns the 'type'
interval from the date.

The following table shows how the @code{type} and @code{expr} arguments
are related:

@multitable @columnfractions .25 .45
@item @code{type} @strong{value} @tab @strong{Expected} @code{expr} @strong{format}
@item @code{SECOND}        @tab @code{SECONDS}
@item @code{MINUTE}        @tab @code{MINUTES}
@item @code{HOUR}          @tab @code{HOURS}
@item @code{DAY}           @tab @code{DAYS}
@item @code{MONTH}         @tab @code{MONTHS}
@item @code{YEAR}          @tab @code{YEARS}
@item @code{MINUTE_SECOND} @tab @code{"MINUTES:SECONDS"}
@item @code{HOUR_MINUTE}   @tab @code{"HOURS:MINUTES"}
@item @code{DAY_HOUR}      @tab @code{"DAYS HOURS"}
@item @code{YEAR_MONTH}    @tab @code{"YEARS-MONTHS"}
@item @code{HOUR_SECOND}   @tab @code{"HOURS:MINUTES:SECONDS"}
@item @code{DAY_MINUTE}    @tab @code{"DAYS HOURS:MINUTES"}
@item @code{DAY_SECOND}    @tab @code{"DAYS HOURS:MINUTES:SECONDS"}
@end multitable

MySQL allows any punctuation delimiter in the @code{expr} format.
Those shown in the table are the suggested delimiters.  If the @code{date}
argument is a @code{DATE} value and your calculations involve only
@code{YEAR}, @code{MONTH}, and @code{DAY} parts (that is, no time parts), the
result is a @code{DATE} value.  Otherwise, the result is a @code{DATETIME}
value:

@example
mysql> SELECT "1997-12-31 23:59:59" + INTERVAL 1 SECOND;
        -> 1998-01-01 00:00:00
mysql> SELECT INTERVAL 1 DAY + "1997-12-31";
        -> 1998-01-01
mysql> SELECT "1998-01-01" - INTERVAL 1 SECOND;
       -> 1997-12-31 23:59:59
mysql> SELECT DATE_ADD("1997-12-31 23:59:59",
    ->                 INTERVAL 1 SECOND);
        -> 1998-01-01 00:00:00
mysql> SELECT DATE_ADD("1997-12-31 23:59:59",
    ->                 INTERVAL 1 DAY);
        -> 1998-01-01 23:59:59
mysql> SELECT DATE_ADD("1997-12-31 23:59:59",
    ->                 INTERVAL "1:1" MINUTE_SECOND);
        -> 1998-01-01 00:01:00
mysql> SELECT DATE_SUB("1998-01-01 00:00:00",
    ->                 INTERVAL "1 1:1:1" DAY_SECOND);
        -> 1997-12-30 22:58:59
mysql> SELECT DATE_ADD("1998-01-01 00:00:00",
    ->                 INTERVAL "-1 10" DAY_HOUR);
        -> 1997-12-30 14:00:00
mysql> SELECT DATE_SUB("1998-01-02", INTERVAL 31 DAY);
        -> 1997-12-02
@end example

If you specify an interval value that is too short (does not include all the
interval parts that would be expected from the @code{type} keyword),
MySQL assumes you have left out the leftmost parts of the interval
value.  For example, if you specify a @code{type} of @code{DAY_SECOND}, the
value of @code{expr} is expected to have days, hours, minutes, and seconds
parts.  If you specify a value like @code{"1:10"}, MySQL assumes
that the days and hours parts are missing and the value represents minutes
and seconds.  In other words, @code{"1:10" DAY_SECOND} is interpreted in such
a way that it is equivalent to @code{"1:10" MINUTE_SECOND}.  This is
analogous to the way that MySQL interprets @code{TIME} values
as representing elapsed time rather than as time of day.

Note that if you add or subtract a date value against something that
contains a time part, the date value will be automatically converted to a
datetime value:

@example
mysql> SELECT DATE_ADD("1999-01-01", INTERVAL 1 DAY);
       -> 1999-01-02
mysql> SELECT DATE_ADD("1999-01-01", INTERVAL 1 HOUR);
       -> 1999-01-01 01:00:00
@end example

If you use really incorrect dates, the result is @code{NULL}. If you add
@code{MONTH}, @code{YEAR_MONTH}, or @code{YEAR} and the resulting date
has a day that is larger than the maximum day for the new month, the day is
adjusted to the maximum days in the new month:

@example
mysql> SELECT DATE_ADD('1998-01-30', INTERVAL 1 MONTH);
        -> 1998-02-28
@end example

Note from the preceding example that the word @code{INTERVAL} and the
@code{type} keyword are not case-sensitive.

@findex EXTRACT()
@item EXTRACT(type FROM date)

The @code{EXTRACT()} function uses the same kinds of interval type
specifiers as @code{DATE_ADD()} or @code{DATE_SUB()}, but extracts parts
from the date rather than performing date arithmetic.

@example
mysql> SELECT EXTRACT(YEAR FROM "1999-07-02");
       -> 1999
mysql> SELECT EXTRACT(YEAR_MONTH FROM "1999-07-02 01:02:03");
       -> 199907
mysql> SELECT EXTRACT(DAY_MINUTE FROM "1999-07-02 01:02:03");
       -> 20102
@end example

@findex TO_DAYS()
@item TO_DAYS(date)
Given a date @code{date}, returns a daynumber (the number of days since year
0):

@example
mysql> SELECT TO_DAYS(950501);
        -> 728779
mysql> SELECT TO_DAYS('1997-10-07');
        -> 729669
@end example

@code{TO_DAYS()} is not intended for use with values that precede the advent
of the Gregorian calendar (1582), because it doesn't take into account the
days that were lost when the calendar was changed.

@findex FROM_DAYS()
@item FROM_DAYS(N)
Given a daynumber @code{N}, returns a @code{DATE} value:

@example
mysql> SELECT FROM_DAYS(729669);
        -> '1997-10-07'
@end example

@code{FROM_DAYS()} is not intended for use with values that precede the
advent of the Gregorian calendar (1582), because it doesn't take into account
the days that were lost when the calendar was changed.

@findex DATE_FORMAT()
@item DATE_FORMAT(date,format)
Formats the @code{date} value according to the @code{format} string. The
following specifiers may be used in the @code{format} string:
@multitable @columnfractions .10 .65
@item @strong{Specifier} @tab @strong{Description}
@item @code{%M} @tab Month name (@code{January}..@code{December})
@item @code{%W} @tab Weekday name (@code{Sunday}..@code{Saturday})
@item @code{%D} @tab Day of the month with English suffix (@code{1st}, @code{2nd}, @code{3rd}, etc.)
@item @code{%Y} @tab Year, numeric, 4 digits
@item @code{%y} @tab Year, numeric, 2 digits
@item @code{%X} @tab Year for the week where Sunday is the first day of the week, numeric, 4 digits, used with '%V'
@item @code{%x} @tab Year for the week, where Monday is the first day of the week, numeric, 4 digits, used with '%v'
@item @code{%a} @tab Abbreviated weekday name (@code{Sun}..@code{Sat})
@item @code{%d} @tab Day of the month, numeric (@code{00}..@code{31})
@item @code{%e} @tab Day of the month, numeric (@code{0}..@code{31})
@item @code{%m} @tab Month, numeric (@code{01}..@code{12})
@item @code{%c} @tab Month, numeric (@code{1}..@code{12})
@item @code{%b} @tab Abbreviated month name (@code{Jan}..@code{Dec})
@item @code{%j} @tab Day of year (@code{001}..@code{366})
@item @code{%H} @tab Hour (@code{00}..@code{23})
@item @code{%k} @tab Hour (@code{0}..@code{23})
@item @code{%h} @tab Hour (@code{01}..@code{12})
@item @code{%I} @tab Hour (@code{01}..@code{12})
@item @code{%l} @tab Hour (@code{1}..@code{12})
@item @code{%i} @tab Minutes, numeric (@code{00}..@code{59})
@item @code{%r} @tab Time, 12-hour (@code{hh:mm:ss [AP]M})
@item @code{%T} @tab Time, 24-hour (@code{hh:mm:ss})
@item @code{%S} @tab Seconds (@code{00}..@code{59})
@item @code{%s} @tab Seconds (@code{00}..@code{59})
@item @code{%p} @tab @code{AM} or @code{PM}
@item @code{%w} @tab Day of the week (@code{0}=Sunday..@code{6}=Saturday)
@item @code{%U} @tab Week (@code{00}..@code{53}), where Sunday is the first day of the week
@item @code{%u} @tab Week (@code{00}..@code{53}), where Monday is the first day of the week
@item @code{%V} @tab Week (@code{01}..@code{53}), where Sunday is the first day of the week. Used with '%X'
@item @code{%v} @tab Week (@code{01}..@code{53}), where Monday is the first day of the week. Used with '%x'
@item @code{%%} @tab A literal @samp{%}.
@end multitable

All other characters are just copied to the result without interpretation:

@example
mysql> SELECT DATE_FORMAT('1997-10-04 22:23:00', '%W %M %Y');
        -> 'Saturday October 1997'
mysql> SELECT DATE_FORMAT('1997-10-04 22:23:00', '%H:%i:%s');
        -> '22:23:00'
mysql> SELECT DATE_FORMAT('1997-10-04 22:23:00',
                          '%D %y %a %d %m %b %j');
        -> '4th 97 Sat 04 10 Oct 277'
mysql> SELECT DATE_FORMAT('1997-10-04 22:23:00',
                          '%H %k %I %r %T %S %w');
        -> '22 22 10 10:23:00 PM 22:23:00 00 6'
mysql> SELECT DATE_FORMAT('1999-01-01', '%X %V');
        -> '1998 52'
@end example

As of MySQL Version 3.23, the @samp{%} character is required before
format specifier characters.  In earlier versions of MySQL,
@samp{%} was optional.

@findex TIME_FORMAT()
@item TIME_FORMAT(time,format)
This is used like the @code{DATE_FORMAT()} function above, but the
@code{format} string may contain only those format specifiers that handle
hours, minutes, and seconds.  Other specifiers produce a @code{NULL} value or
@code{0}.

@findex CURDATE()
@findex CURRENT_DATE
@item CURDATE()
@itemx CURRENT_DATE
Returns today's date as a value in @code{'YYYY-MM-DD'} or @code{YYYYMMDD}
format, depending on whether the function is used in a string or numeric
context:

@example
mysql> SELECT CURDATE();
        -> '1997-12-15'
mysql> SELECT CURDATE() + 0;
        -> 19971215
@end example

@findex CURTIME()
@findex CURRENT_TIME
@item CURTIME()
@itemx CURRENT_TIME
Returns the current time as a value in @code{'HH:MM:SS'} or @code{HHMMSS}
format, depending on whether the function is used in a string or numeric
context:

@example
mysql> SELECT CURTIME();
        -> '23:50:26'
mysql> SELECT CURTIME() + 0;
        -> 235026
@end example

@findex NOW()
@findex SYSDATE()
@findex CURRENT_TIMESTAMP
@item NOW()
@itemx SYSDATE()
@itemx CURRENT_TIMESTAMP
Returns the current date and time as a value in @code{'YYYY-MM-DD HH:MM:SS'}
or @code{YYYYMMDDHHMMSS} format, depending on whether the function is used in
a string or numeric context:

@example
mysql> SELECT NOW();
        -> '1997-12-15 23:50:26'
mysql> SELECT NOW() + 0;
        -> 19971215235026
@end example

@findex UNIX_TIMESTAMP()
@item UNIX_TIMESTAMP()
@itemx UNIX_TIMESTAMP(date)
If called with no argument, returns a Unix timestamp (seconds since
@code{'1970-01-01 00:00:00'} GMT) as an unsigned integer. If
@code{UNIX_TIMESTAMP()} is called with a @code{date} argument, it
returns the value of the argument as seconds since @code{'1970-01-01
00:00:00'} GMT.  @code{date} may be a @code{DATE} string, a
@code{DATETIME} string, a @code{TIMESTAMP}, or a number in the format
@code{YYMMDD} or @code{YYYYMMDD} in local time:

@example
mysql> SELECT UNIX_TIMESTAMP();
        -> 882226357
mysql> SELECT UNIX_TIMESTAMP('1997-10-04 22:23:00');
        -> 875996580
@end example

When @code{UNIX_TIMESTAMP} is used on a @code{TIMESTAMP} column, the function
will return the internal timestamp value directly, with no implicit
``string-to-unix-timestamp'' conversion.
If you pass an out-of-range date to @code{UNIX_TIMESTAMP()} it will
return 0, but please note that only basic checking is performed
(year 1970-2037, month 01-12, day 01-31).

If you want to subtract @code{UNIX_TIMESTAMP()} columns, you may want to
cast the result to signed integers. @xref{Cast Functions}.

@findex FROM_UNIXTIME()
@item FROM_UNIXTIME(unix_timestamp)
Returns a representation of the @code{unix_timestamp} argument as a value in
@code{'YYYY-MM-DD HH:MM:SS'} or @code{YYYYMMDDHHMMSS} format, depending on
whether the function is used in a string or numeric context:

@example
mysql> SELECT FROM_UNIXTIME(875996580);
        -> '1997-10-04 22:23:00'
mysql> SELECT FROM_UNIXTIME(875996580) + 0;
        -> 19971004222300
@end example

@findex FROM_UNIXTIME()
@item FROM_UNIXTIME(unix_timestamp,format)
Returns a string representation of the Unix timestamp, formatted according to
the @code{format} string. @code{format} may contain the same specifiers as
those listed in the entry for the @code{DATE_FORMAT()} function:

@example
mysql> SELECT FROM_UNIXTIME(UNIX_TIMESTAMP(),
                            '%Y %D %M %h:%i:%s %x');
        -> '1997 23rd December 03:43:30 1997'
@end example

@findex SEC_TO_TIME()
@item SEC_TO_TIME(seconds)
Returns the @code{seconds} argument, converted to hours, minutes, and seconds,
as a value in @code{'HH:MM:SS'} or @code{HHMMSS} format, depending on whether
the function is used in a string or numeric context:

@example
mysql> SELECT SEC_TO_TIME(2378);
        -> '00:39:38'
mysql> SELECT SEC_TO_TIME(2378) + 0;
        -> 3938
@end example

@findex TIME_TO_SEC()
@item TIME_TO_SEC(time)
Returns the @code{time} argument, converted to seconds:

@example
mysql> SELECT TIME_TO_SEC('22:23:00');
        -> 80580
mysql> SELECT TIME_TO_SEC('00:39:38');
        -> 2378
@end example
@end table

@node Cast Functions, Other Functions, Date and time functions, Functions
@subsection Cast Functions

The syntax of the @code{CAST} function is:

@findex CAST
@findex CONVERT

@example
CAST(expression AS type)

or

CONVERT(expression,type)
@end example

Where type is one of:
@itemize @bullet
@item
@code{BINARY}
@item
@code{DATE}
@item
@code{DATETIME}
@item
@code{SIGNED @{INTEGER@}}
@item
@code{TIME}
@item
@code{UNSIGNED @{INTEGER@}}
@end itemize

@code{CAST()} is ANSI SQL99 syntax and @code{CONVERT()} is ODBC syntax.

The cast function is mainly useful when you want to create a column with
a specific type in a @code{CREATE ... SELECT}:

@example
CREATE TABLE new_table SELECT CAST('2000-01-01' AS DATE);
@end example

@code{CAST(string AS BINARY} is the same thing as @code{BINARY string}.

To cast a string to a numeric value, you don't normally have to do
anything; just use the string value as it would be a number:

@example
mysql> SELECT 1+'1';
        -> 2
@end example

MySQL supports arithmetic with both signed and unsigned 64-bit values.
If you are using an numerical operations (like @code{+}) and one of the
operands are @code{unsigned integer}, then the result will be unsigned.
You can override this by using the @code{SIGNED} and @code{UNSIGNED}
cast operators, which will cast the operation to a signed or
unsigned 64-bit integer, respectively.

@example
mysql> SELECT CAST(1-2 AS UNSIGNED)
        -> 18446744073709551615
mysql> SELECT CAST(CAST(1-2 AS UNSIGNED) AS SIGNED);
        -> -1
@end example

Note that if either operation is a floating-point value (In this context
@code{DECIMAL()} is regarded as a floating-point value) the result will
be a floating-point value and is not affected by the above rule.

@example
mysql> SELECT CAST(1 AS UNSIGNED) -2.0
        -> -1.0
@end example

If you are using a string in an arithmetic operation, this is converted
to a floating-point number.

The @code{CAST()} and @code{CONVERT()} functions were added in MySQL 4.0.2.

The handing of unsigned values was changed in MySQL 4.0 to be able to
support @code{BIGINT} values properly. If you have some code that you
want to run in both MySQL 4.0 and 3.23 (in which case you probably can't
use the CAST function), you can use the following trick to get a signed
result when subtracting two unsigned integer columns:

@example
SELECT (unsigned_column_1+0.0)-(unsigned_column_2+0.0);
@end example

The idea is that the columns are converted to floating-point before doing
the subtraction.

If you get a problem with @code{UNSIGNED} columns in your old MySQL
application when porting to MySQL 4.0, you can use the
@code{--sql-mode=NO_UNSIGNED_SUBTRACTION} option when starting
@code{mysqld}.  Note however that as long as you use this, you will not
be able to make efficient use of the @code{UNSIGNED BIGINT} column type.

@node Other Functions, Group by functions, Cast Functions, Functions
@subsection Other Functions

@menu
* Bit functions::               Bit Functions
* Miscellaneous functions::     Miscellaneous Functions
@end menu


@node Bit functions, Miscellaneous functions, Other Functions, Other Functions
@subsubsection Bit Functions

@findex arithmetic functions
@findex bit functions
@findex functions, arithmetic
@findex functions, bit

MySQL uses @code{BIGINT} (64-bit) arithmetic for bit operations, so
these operators have a maximum range of 64 bits.

@table @code
@findex | (bitwise OR)
@findex OR, bitwise
@item |
Bitwise OR
@example
mysql> SELECT 29 | 15;
        -> 31
@end example

The result is an unsigned 64-bit integer.

@findex & (bitwise AND)
@findex AND, bitwise
@item &
Bitwise AND:
@example
mysql> SELECT 29 & 15;
        -> 13
@end example

The result is an unsigned 64-bit integer.

@findex << (left shift)
@item <<
Shifts a longlong (@code{BIGINT}) number to the left:
@example
mysql> SELECT 1 << 2;
        -> 4
@end example

The result is an unsigned 64-bit integer.

@findex >> (right shift)
@item >>
Shifts a longlong (@code{BIGINT}) number to the right:
@example
mysql> SELECT 4 >> 2;
        -> 1
@end example

The result is an unsigned 64-bit integer.

@findex ~
@item ~
Invert all bits:
@example
mysql> SELECT 5 & ~1;
        -> 4
@end example

The result is an unsigned 64-bit integer.

@findex BIT_COUNT()
@item BIT_COUNT(N)
Returns the number of bits that are set in the argument @code{N}:
@example
mysql> SELECT BIT_COUNT(29);
        -> 4
@end example
@end table


@node Miscellaneous functions,  , Bit functions, Other Functions
@subsubsection Miscellaneous Functions

@findex miscellaneous functions
@findex functions, miscellaneous

@table @code
@findex DATABASE()
@item DATABASE()
Returns the current database name:

@example
mysql> SELECT DATABASE();
        -> 'test'
@end example

If there is no current database, @code{DATABASE()} returns the empty string.

@findex USER()
@findex SYSTEM_USER()
@findex SESSION_USER()
@item USER()
@itemx SYSTEM_USER()
@itemx SESSION_USER()
Returns the current MySQL user name:

@example
mysql> SELECT USER();
        -> 'davida@@localhost'
@end example

In MySQL Version 3.22.11 or later, this includes the client hostname
as well  as the user name.  You can extract just the user name part like this
(which works whether the value includes a hostname part):

@example
mysql> SELECT SUBSTRING_INDEX(USER(),"@@",1);
        -> 'davida'
@end example

@findex PASSWORD()
@item PASSWORD(str)
Calculates a password string from the plaintext password @code{str}. This is
the function that is used for encrypting MySQL passwords for storage
in the @code{Password} column of the @code{user} grant table:

@example
mysql> SELECT PASSWORD('badpwd');
        -> '7f84554057dd964b'
@end example

@cindex password encryption, reversibility of
@code{PASSWORD()} encryption is non-reversible.

@code{PASSWORD()} does not perform password encryption in the same way that
Unix passwords are encrypted.  You should not assume that if your Unix
password and your MySQL password are the same, @code{PASSWORD()}
will result in the same encrypted value as is stored in the Unix password
file.  See @code{ENCRYPT()}.

@findex ENCRYPT()
@item ENCRYPT(str[,salt])
Encrypt @code{str} using the Unix @code{crypt()} system call. The
@code{salt} argument should be a string with two characters.
(As of MySQL Version 3.22.16, @code{salt} may be longer than two characters.):

@example
mysql> SELECT ENCRYPT("hello");
        -> 'VxuFAJXVARROc'
@end example

If @code{crypt()} is not available on your system, @code{ENCRYPT()} always
returns @code{NULL}.

@code{ENCRYPT()} ignores all but the first 8 characters of @code{str}, at
least on some systems.  This will be determined by the behavior of the
underlying @code{crypt()} system call.

@findex ENCODE()
@item ENCODE(str,pass_str)
Encrypt @code{str} using @code{pass_str} as the password.
To decrypt the result, use @code{DECODE()}.

The results is a binary string of the same length as @code{string}.
If you want to save it in a column, use a @code{BLOB} column type.

@findex DECODE()
@item DECODE(crypt_str,pass_str)
Descrypts the encrypted string @code{crypt_str} using @code{pass_str} as the
password.  @code{crypt_str} should be a string returned from
@code{ENCODE()}.

@findex MD5()
@item MD5(string)
Calculates a MD5 checksum for the string. Value is returned as a 32 long
hex number that may, for example, be used as a hash key:

@example
mysql> SELECT MD5("testing");
        -> 'ae2b1fca515949e5d54fb22b8ed95575'
@end example

This is an "RSA Data Security, Inc. MD5 Message-Digest Algorithm".

@findex DES_ENCRYPT()
@item DES_ENCRYPT(string_to_encrypt [, (key_number | key_string) ] )

Encrypts the string with the given key using the DES algorithm, which
provides strong encryption.

Note that this function only works if you have configured MySQL with
SSL support. @xref{Secure connections}.

The encryption key to use is chosen the following way:

@multitable @columnfractions .25 .65
@item @strong{Argument} @tab @strong{Description}
@item Only one argument @tab
The first key from @code{des-key-file} is used.
@item key number @tab
The given key (0-9) from the @code{des-key-file} is used.
@item string @tab
The given @code{key_string} will be used to crypt @code{string_to_encrypt}.
@end multitable

The return string will be a binary string where the first character
will be @code{CHAR(128 | key_number)}.

The 128 is added to make it easier to recognize an encrypted key.
If you use a string key, @code{key_number} will be 127.

On error, this function returns @code{NULL}.

The string length for the result will be
@code{new_length= org_length + (8-(org_length % 8))+1}.

The @code{des-key-file} has the following format:

@example
key_number des_key_string
key_number des_key_string
@end example

Each @code{key_number} must be a number in the range from 0 to 9.  Lines in
the file may be in any order.  @code{des_key_string} is the string that
will be used to encrypt the message.  Between the number and the key there
should be at least one space.  The first key is the default key that will
be used if you don't specify any key argument to @code{DES_ENCRYPT()}

You can tell MySQL to read new key values from the key file with the
@code{FLUSH DES_KEY_FILE} command.  This requires the @code{Reload_priv}
privilege.

One benefit of having a set of default keys is that it gives applications
a way to check for existence of encrypted column values, without giving
the end user the right to decrypt those values.

@example
mysql> SELECT customer_address FROM customer_table WHERE
       crypted_credit_card = DES_ENCRYPT("credit_card_number");
@end example

@findex DES_DECRYPT()
@item DES_DECRYPT(string_to_decrypt [, key_string])

Decrypts a string encrypted with @code{DES_ENCRYPT()}.

Note that this function only works if you have configured MySQL with
SSL support. @xref{Secure connections}.

If no @code{key_string} argument is given, @code{DES_DECRYPT()} examines
the first byte of the encrypted string to determine the DES key number
that was used to encrypt the original string, then reads the key
from the @code{des-key-file} to decrypt the message.  For this to work
the user must have the @strong{process} privilege.

If you pass this function a @code{key_string} argument, that string
is used as the key for decrypting the message.

If the @code{string_to_decrypt} doesn't look like an encrypted string, MySQL
will return the given @code{string_to_decrypt}.

On error, this function returns @code{NULL}.

@findex LAST_INSERT_ID([expr])
@item LAST_INSERT_ID([expr])
Returns the last automatically generated value that was inserted into an
@code{AUTO_INCREMENT} column.
@xref{mysql_insert_id,, @code{mysql_insert_id()}}.

@example
mysql> SELECT LAST_INSERT_ID();
        -> 195
@end example

The last ID that was generated is maintained in the server on a
per-connection basis.  It will not be changed by another client. It will not
even be changed if you update another @code{AUTO_INCREMENT} column with a
non-magic value (that is, a value that is not @code{NULL} and not @code{0}).

If you insert many rows at the same time with an insert statement,
@code{LAST_INSERT_ID()} returns the value for the first inserted row.
The reason for this is so that you it makes it possible to easily reproduce
the same @code{INSERT} statement against some other server.

@cindex sequence emulation
If @code{expr} is given as an argument to @code{LAST_INSERT_ID()}, then
the value of the argument is returned by the function, is set as the
next value to be returned by @code{LAST_INSERT_ID()} and used as the next
@code{AUTO_INCREMENT} value.  This can be used to simulate sequences:

First create the table:

@example
mysql> CREATE TABLE sequence (id INT NOT NULL);
mysql> INSERT INTO sequence VALUES (0);
@end example

Then the table can be used to generate sequence numbers like this:

@example
mysql> UPDATE sequence SET id=LAST_INSERT_ID(id+1);
@end example

You can generate sequences without calling @code{LAST_INSERT_ID()}, but the
utility of using the function this way is that the ID value is maintained in
the server as the last automatically generated value.  You can retrieve the
new ID as you would read any normal @code{AUTO_INCREMENT} value in
MySQL.  For example, @code{LAST_INSERT_ID()} (without an argument)
will return the new ID.  The C API function @code{mysql_insert_id()}
can also be used to get the value.

Note that as @code{mysql_insert_id()} is only updated after
@code{INSERT} and @code{UPDATE} statements, you can't use this function
to retrieve the value for @code{LAST_INSERT_ID(expr)} after executing
other SQL statements like @code{SELECT} or @code{SET}.


@findex FORMAT()
@item FORMAT(X,D)
Formats the number @code{X} to a format like @code{'#,###,###.##'}, rounded
to @code{D} decimals.  If @code{D} is @code{0}, the result will have no
decimal point or fractional part:

@example
mysql> SELECT FORMAT(12332.123456, 4);
        -> '12,332.1235'
mysql> SELECT FORMAT(12332.1,4);
        -> '12,332.1000'
mysql> SELECT FORMAT(12332.2,0);
        -> '12,332'
@end example

@findex VERSION()
@item VERSION()
Returns a string indicating the MySQL server version:

@example
mysql> SELECT VERSION();
        -> '3.23.13-log'
@end example

Note that if your version ends with @code{-log} this means that logging is
enabled.

@findex CONNECTION_ID()
@item CONNECTION_ID()
Returns the connection id (@code{thread_id}) for the connection.
Every connection has its own unique id:

@example
mysql> SELECT CONNECTION_ID();
        -> 1
@end example

@cindex timeout
@findex GET_LOCK()
@item GET_LOCK(str,timeout)
Tries to obtain a lock with a name given by the string @code{str}, with a
timeout of @code{timeout} seconds.  Returns @code{1} if the lock was obtained
successfully, @code{0} if the attempt timed out, or @code{NULL} if an error
occurred (such as running out of memory or the thread was killed with
@code{mysqladmin kill}).  A lock is released when you execute
@code{RELEASE_LOCK()}, execute a new @code{GET_LOCK()}, or the thread
terminates.  This function can be used to implement application locks or to
simulate record locks.  It blocks requests by other clients for locks with
the same name; clients that agree on a given lock string name can use the
string to perform cooperative advisory locking:

@example
mysql> SELECT GET_LOCK("lock1",10);
        -> 1
mysql> SELECT GET_LOCK("lock2",10);
        -> 1
mysql> SELECT RELEASE_LOCK("lock2");
        -> 1
mysql> SELECT RELEASE_LOCK("lock1");
        -> NULL
@end example

Note that the second @code{RELEASE_LOCK()} call returns @code{NULL} because
the lock @code{"lock1"} was automatically released by the second
@code{GET_LOCK()} call.

@findex RELEASE_LOCK()
@item RELEASE_LOCK(str)
Releases the lock named by the string @code{str} that was obtained with
@code{GET_LOCK()}. Returns @code{1} if the lock was released, @code{0} if the
lock wasn't locked by this thread (in which case the lock is not released),
and @code{NULL} if the named lock didn't exist.  The lock will not exist if
it was never obtained by a call to @code{GET_LOCK()} or if it already has
been released.

The @code{DO} statement is convinient to use with @code{RELEASE_LOCK()}.
@xref{DO}.

@findex BENCHMARK()
@item BENCHMARK(count,expr)
The @code{BENCHMARK()} function executes the expression @code{expr}
repeatedly @code{count} times.  It may be used to time how fast MySQL
processes the expression.  The result value is always @code{0}.  The intended
use is in the @code{mysql} client, which reports query execution times:

@example
mysql> SELECT BENCHMARK(1000000,ENCODE("hello","goodbye"));
+----------------------------------------------+
| BENCHMARK(1000000,ENCODE("hello","goodbye")) |
+----------------------------------------------+
|                                            0 |
+----------------------------------------------+
1 row in set (4.74 sec)
@end example

The time reported is elapsed time on the client end, not CPU time on the
server end.  It may be advisable to execute @code{BENCHMARK()} several
times, and interpret the result with regard to how heavily loaded the
server machine is.

@findex INET_NTOA()
@item INET_NTOA(expr)
Given a numeric network address (4 or 8 byte), returns the dotted-quad
representation of the address as a string:

@example
mysql> SELECT INET_NTOA(3520061480);
       ->  "209.207.224.40"
@end example

@findex INET_ATON()
@item INET_ATON(expr)
Given the dotted-quad representation of a network address as a string,
returns an integer that represents the numeric value of the address.
Addresses may be 4 or 8 byte addresses:

@example
mysql> SELECT INET_ATON("209.207.224.40");
       ->  3520061480
@end example

The generated number is always in network byte order; for example the
above number is calculated as @code{209*256^3 + 207*256^2 + 224*256 +40}.

@findex MASTER_POS_WAIT()
@item MASTER_POS_WAIT(log_name, log_pos)
Blocks until the slave reaches the specified position in the master
log during replication. If master information is not initialised,
returns @code{NULL}. If the slave is not running, will block and wait
until it is started and goes to or past the specified position. If the
slave is already past the specified position, returns immediately. The
return value is the number of log events it had to wait to get to the
specified position, or @code{NULL} in case of error. Useful for control
of master-slave synchronisation, but was originally written to facilitate
replication testing.

@findex FOUND_ROWS()
@findex LIMIT
@item FOUND_ROWS()
Returns the number of rows that the last @code{SELECT SQL_CALC_FOUND_ROWS ...}
command would have returned, if wasn't restricted with @code{LIMIT}.

@example
mysql> SELECT SQL_CALC_FOUND_ROWS * FROM tbl_name
       WHERE id > 100 LIMIT 10;
mysql> SELECT FOUND_ROWS();
@end example

The second @code{SELECT} will return a number indicating how many rows the
first @code{SELECT} would have returned had it been written without the
@code{LIMIT} clause.

Note that if you are using @code{SELECT SQL_CALC_FOUND_ROWS ...} MySQL has
to calculate all rows in the result set.  However, this is faster than
if you would not use @code{LIMIT}, as the result set need not be sent
to the client.
@end table


@node Group by functions,  , Other Functions, Functions
@subsection Functions for Use with @code{GROUP BY} Clauses

@findex GROUP BY functions
@findex functions, GROUP BY

If you use a group function in a statement containing no @code{GROUP BY}
clause, it is equivalent to grouping on all rows.

@table @code
@findex COUNT()
@item COUNT(expr)
Returns a count of the number of non-@code{NULL} values in the rows
retrieved by a @code{SELECT} statement:

@example
mysql> SELECT student.student_name,COUNT(*)
    ->        FROM student,course
    ->        WHERE student.student_id=course.student_id
    ->        GROUP BY student_name;

@end example

@code{COUNT(*)} is somewhat different in that it returns a count of
the number of rows retrieved, whether they contain @code{NULL}
values.

@code{COUNT(*)} is optimised to
return very quickly if the @code{SELECT} retrieves from one table, no
other columns are retrieved, and there is no @code{WHERE} clause.
For example:

@example
mysql> SELECT COUNT(*) FROM student;
@end example

@findex COUNT(DISTINCT)
@findex DISTINCT
@item COUNT(DISTINCT expr,[expr...])
Returns a count of the number of different non-@code{NULL} values:

@example
mysql> SELECT COUNT(DISTINCT results) FROM student;
@end example

In MySQL you can get the number of distinct expression
combinations that don't contain NULL by giving a list of expressions.
In ANSI SQL you would have to do a concatenation of all expressions
inside @code{CODE(DISTINCT ...)}.

@findex AVG()
@item AVG(expr)
Returns the average value of @code{expr}:

@example
mysql> SELECT student_name, AVG(test_score)
    ->        FROM student
    ->        GROUP BY student_name;
@end example

@findex MIN()
@findex MAX()
@item MIN(expr)
@itemx MAX(expr)
Returns the minimum or maximum value of @code{expr}.  @code{MIN()} and
@code{MAX()} may take a string argument; in such cases they return the
minimum or maximum string value. @xref{MySQL indexes}.

@example
mysql> SELECT student_name, MIN(test_score), MAX(test_score)
    ->        FROM student
    ->        GROUP BY student_name;
@end example

@findex SUM()
@item SUM(expr)
Returns the sum of @code{expr}.  Note that if the return set has no rows,
it returns NULL!

@findex STD()
@findex STDDEV()
@cindex Oracle compatibility
@cindex compatibility, with Oracle
@item STD(expr)
@itemx STDDEV(expr)
Returns the standard deviation of @code{expr}. This is an extension to
ANSI SQL. The @code{STDDEV()} form of this function is provided for Oracle
compatibility.

@findex BIT_OR()
@item BIT_OR(expr)
Returns the bitwise @code{OR} of all bits in @code{expr}. The calculation is
performed with 64-bit (@code{BIGINT}) precision.

@findex BIT_AND()
@item BIT_AND(expr)
Returns the bitwise @code{AND} of all bits in @code{expr}. The calculation is
performed with 64-bit (@code{BIGINT}) precision.
@end table

@cindex @code{GROUP BY}, extensions to ANSI SQL
MySQL has extended the use of @code{GROUP BY}. You can use columns or
calculations in the @code{SELECT} expressions that don't appear in
the @code{GROUP BY} part. This stands for @emph{any possible value for this
group}.  You can use this to get better performance by avoiding sorting and
grouping on unnecessary items.  For example, you don't need to group on
@code{customer.name} in the following query:

@example
mysql> SELECT order.custid,customer.name,MAX(payments)
    ->        FROM order,customer
    ->        WHERE order.custid = customer.custid
    ->        GROUP BY order.custid;
@end example

In ANSI SQL, you would have to add @code{customer.name} to the @code{GROUP
BY} clause.  In MySQL, the name is redundant if you don't run in
ANSI mode.

@strong{Don't use this feature} if the columns you omit from the
@code{GROUP BY} part aren't unique in the group!  You will get
unpredictable results.

In some cases, you can use @code{MIN()} and @code{MAX()} to obtain a specific
column value even if it isn't unique. The following gives the value of
@code{column} from the row containing the smallest value in the @code{sort}
column:

@example
SUBSTR(MIN(CONCAT(RPAD(sort,6,' '),column)),7)
@end example

@xref{example-Maximum-column-group-row}.

@cindex @code{ORDER BY}, aliases in
@cindex aliases, in @code{ORDER BY} clauses
@cindex @code{GROUP BY}, aliases in
@cindex aliases, in @code{GROUP BY} clauses
@cindex expression aliases
@cindex aliases, for expressions
Note that if you are using MySQL Version 3.22 (or earlier) or if
you are trying to follow ANSI SQL, you can't use expressions in @code{GROUP
BY} or @code{ORDER BY} clauses.  You can work around this limitation by
using an alias for the expression:

@example
mysql> SELECT id,FLOOR(value/100) AS val FROM tbl_name
    ->        GROUP BY id,val ORDER BY val;
@end example

In MySQL Version 3.23 you can do:

@example
mysql> SELECT id,FLOOR(value/100) FROM tbl_name ORDER BY RAND();
@end example



@node Data Manipulation, Data Definition, Functions, Reference
@section Data Manipulation: @code{SELECT}, @code{INSERT}, @code{UPDATE}, @code{DELETE}


@menu
* SELECT::                      @code{SELECT} Syntax
* HANDLER::                     @code{HANDLER} Syntax
* INSERT::                      @code{INSERT} Syntax
* INSERT DELAYED::              @code{INSERT DELAYED} Syntax
* UPDATE::                      @code{UPDATE} Syntax
* DELETE::                      @code{DELETE} Syntax
* TRUNCATE::                    @code{TRUNCATE} Syntax
* REPLACE::                     @code{REPLACE} Syntax
* LOAD DATA::                   @code{LOAD DATA INFILE} Syntax
* DO::                          @code{DO} Syntax
@end menu

@node SELECT, HANDLER, Data Manipulation, Data Manipulation
@subsection @code{SELECT} Syntax

@findex SELECT

@c help SELECT
@example
SELECT [STRAIGHT_JOIN]
       [SQL_SMALL_RESULT] [SQL_BIG_RESULT] [SQL_BUFFER_RESULT]
       [SQL_CACHE | SQL_NO_CACHE] [SQL_CALC_FOUND_ROWS] [HIGH_PRIORITY]
       [DISTINCT | DISTINCTROW | ALL]
    select_expression,...
    [INTO @{OUTFILE | DUMPFILE@} 'file_name' export_options]
    [FROM table_references
      [WHERE where_definition]
      [GROUP BY @{unsigned_integer | col_name | formula@} [ASC | DESC], ...]
      [HAVING where_definition]
      [ORDER BY @{unsigned_integer | col_name | formula@} [ASC | DESC] ,...]
      [LIMIT [offset,] rows]
      [PROCEDURE procedure_name]
      [FOR UPDATE | LOCK IN SHARE MODE]]
@end example
@c help end

@code{SELECT} is used to retrieve rows selected from one or more tables.
@code{select_expression} indicates the columns you want to retrieve.
@code{SELECT} may also be used to retrieve rows computed without reference to
any table.  For example:

@example
mysql> SELECT 1 + 1;
         -> 2
@end example

All keywords used must be given in exactly the order shown above. For example,
a @code{HAVING} clause must come after any @code{GROUP BY} clause and before
any @code{ORDER BY} clause.

@itemize @bullet

@item
@cindex aliases, on expressions
@cindex expression aliases
A @code{SELECT} expression may be given an alias using @code{AS}. The alias
is used as the expression's column name and can be used with
@code{ORDER BY} or @code{HAVING} clauses.  For example:

@example
mysql> SELECT CONCAT(last_name,', ',first_name) AS full_name
    FROM mytable ORDER BY full_name;
@end example

@item
It is not allowed to use a column alias in a @code{WHERE} clause,
because the column value may not yet be determined when the
@code{WHERE} clause is executed.
@xref{Problems with alias}.

@item
@findex AS
@findex USE INDEX
@findex IGNORE INDEX
@findex USE KEY
@findex IGNORE KEY
The @code{FROM table_references} clause indicates the tables from which to
retrieve rows.  If you name more than one table, you are performing a
join.  For information on join syntax, see @ref{JOIN, , @code{JOIN}}.
For each table specified, you may optionally specify an alias.
@example
table_name [[AS] alias] [USE INDEX (key_list)] [IGNORE INDEX (key_list)]
@end example
As of MySQL Version 3.23.12, you can give hints about which
index MySQL should use when retrieving information from a
table. This is useful if @code{EXPLAIN} shows that MySQL is
using the wrong index.  By specifying @code{USE INDEX (key_list)}, you
can tell MySQL to use only one of the specified indexes to
find rows in the table.  The alternative syntax @code{IGNORE INDEX
(key_list)} can be used to tell MySQL to not use some
particular index.
@code{USE/IGNORE KEY} are synonyms for @code{USE/IGNORE INDEX}.

@item
You can refer to a column as @code{col_name}, @code{tbl_name.col_name}, or
@code{db_name.tbl_name.col_name}.  You need not specify a @code{tbl_name} or
@code{db_name.tbl_name} prefix for a column reference in a @code{SELECT}
statement unless the reference would be ambiguous.  See @ref{Legal names},
for examples of ambiguity that require the more explicit column reference
forms.

@item
@cindex aliases, for tables
@cindex table aliases
A table reference may be aliased using @code{tbl_name [AS] alias_name}:

@example
mysql> SELECT t1.name, t2.salary FROM employee AS t1, info AS t2
    ->        WHERE t1.name = t2.name;
mysql> SELECT t1.name, t2.salary FROM employee t1, info t2
    ->        WHERE t1.name = t2.name;
@end example

@item
Columns selected for output may be referred to in @code{ORDER BY} and
@code{GROUP BY} clauses using column names, column aliases, or column
positions.  Column positions begin with 1:

@example
mysql> SELECT college, region, seed FROM tournament
    ->        ORDER BY region, seed;
mysql> SELECT college, region AS r, seed AS s FROM tournament
    ->        ORDER BY r, s;
mysql> SELECT college, region, seed FROM tournament
    ->        ORDER BY 2, 3;
@end example

To sort in reverse order, add the @code{DESC} (descending) keyword to the
name of the column in the @code{ORDER BY} clause that you are sorting by.
The default is ascending order; this may be specified explicitly using
the @code{ASC} keyword.

@item
You can in the @code{WHERE} clause use any of the functions that
MySQL support. @xref{Functions}.

@item
The @code{HAVING} clause can refer to any column or alias named in the
@code{select_expression}. It is applied last, just before items are sent to
the client, with no optimisation. Don't use @code{HAVING} for items that
should be in the @code{WHERE} clause.  For example, do not write this:

@example
mysql> SELECT col_name FROM tbl_name HAVING col_name > 0;
@end example

Write this instead:

@example
mysql> SELECT col_name FROM tbl_name WHERE col_name > 0;
@end example

In MySQL Version 3.22.5 or later, you can also write queries like this:

@example
mysql> SELECT user,MAX(salary) FROM users
    ->        GROUP BY user HAVING MAX(salary)>10;
@end example

In older MySQL versions, you can write this instead:

@example
mysql> SELECT user,MAX(salary) AS sum FROM users
    ->        group by user HAVING sum>10;
@end example

@item
The options @code{DISTINCT}, @code{DISTINCTROW} and @code{ALL} specify
whether duplicate rows should be returned. The default is (@code{ALL}),
all matching rows are returned. @code{DISTINCT} and @code{DISTINCTROW}
are synonyms and specify that duplicate rows in the result set should
be removed.

@item
All options beginning with @code{SQL_}, @code{STRAIGHT_JOIN}, and
@code{HIGH_PRIORITY} are MySQL extensions to ANSI SQL.

@item
@code{HIGH_PRIORITY} will give the @code{SELECT} higher priority than
a statement that updates a table.  You should only use this for queries
that are very fast and must be done at once.  A @code{SELECT HIGH_PRIORITY}
query will run if the table is locked for read even if there is an update
statement that is waiting for the table to be free.

@item
@code{SQL_BIG_RESULT} can be used with @code{GROUP BY} or @code{DISTINCT}
to tell the optimiser that the result set will have many rows.  In this case,
MySQL will directly use disk-based temporary tables if needed.
MySQL will also, in this case, prefer sorting to doing a
temporary table with a key on the @code{GROUP BY} elements.

@item
@code{SQL_BUFFER_RESULT} will force the result to be put into a temporary
table. This will help MySQL free the table locks early and will help
in cases where it takes a long time to send the result set to the client.

@item
@code{SQL_SMALL_RESULT}, a MySQL-specific option, can be used
with @code{GROUP BY} or @code{DISTINCT} to tell the optimiser that the
result set will be small.  In this case, MySQL will use fast
temporary tables to store the resulting table instead of using sorting. In
MySQL Version 3.23 this shouldn't normally be needed.

@item
@code{SQL_CALC_FOUND_ROWS} tells MySQL to calculate how many rows there
would be in the result, disregarding any @code{LIMIT} clause. The number
of rows can be obtained with @code{SELECT
FOUND_ROWS()}. @xref{Miscellaneous functions}.

@item
@code{SQL_CACHE} tells MySQL to store the query result in the query cache
if you are using @code{SQL_QUERY_CACHE_TYPE=2} (@code{DEMAND}).
@xref{Query Cache}.

@item
@code{SQL_NO_CACHE} tells MySQL to not allow the query result to be stored
in the query cache. @xref{Query Cache}.

@item
@cindex @code{GROUP BY}, extensions to ANSI SQL
If you use @code{GROUP BY}, the output rows will be sorted according to the
@code{GROUP BY} as if you would have had an @code{ORDER BY} over all the fields
in the @code{GROUP BY}. MySQL has extended the @code{GROUP BY} so that
you can also specify @code{ASC} and @code{DESC} to @code{GROUP BY}:

@example
SELECT a,COUNT(b) FROM test_table GROUP BY a DESC
@end example

@item
MySQL has extended the use of @code{GROUP BY} to allow you to
select fields which are not mentioned in the @code{GROUP BY} clause.
If you are not getting the results you expect from your query, please
read the @code{GROUP BY} description.
@xref{Group by functions}.

@item
@cindex hints
@code{STRAIGHT_JOIN} forces the optimiser to join the tables in the order in
which they are listed in the @code{FROM} clause. You can use this to speed up
a query if the optimiser joins the tables in non-optimal order.
@xref{EXPLAIN, , @code{EXPLAIN}}.

@item
The @code{LIMIT} clause can be used to constrain the number of rows returned
by the @code{SELECT} statement.  @code{LIMIT} takes one or two numeric
arguments.  The arguments must be integer constants.

If two arguments are given, the first specifies the offset of the first row to
return, the second specifies the maximum number of rows to return.
The offset of the initial row is 0 (not 1):

@example
mysql> SELECT * FROM table LIMIT 5,10;  # Retrieve rows 6-15
@end example

If one argument is given, it indicates the maximum number of rows to return:

@example
mysql> SELECT * FROM table LIMIT 5;     # Retrieve first 5 rows
@end example

In other words, @code{LIMIT n} is equivalent to @code{LIMIT 0,n}.

@item
@tindex /etc/passwd
The @code{SELECT ... INTO OUTFILE 'file_name'} form of @code{SELECT} writes
the selected rows to a file. The file is created on the server host and
cannot already exist (among other things, this prevents database tables and
files such as @file{/etc/passwd} from being destroyed).  You must have the
@strong{file} privilege on the server host to use this form of @code{SELECT}.

@code{SELECT ... INTO OUTFILE} is mainly intended to let you very
quickly dump a table on the server machine. If you want to create the
resulting file on some other host than the server host you can't use
@code{SELECT ... INTO OUTFILE}. In this case you should instead use some
client program like @code{mysqldump --tab} or @code{mysql -e "SELECT
..." > outfile} to generate the file.

@code{SELECT ...  INTO OUTFILE} is the complement of @code{LOAD DATA
INFILE}; the syntax for the @code{export_options} part of the statement
consists of the same @code{FIELDS} and @code{LINES} clauses that are used
with the @code{LOAD DATA INFILE} statement.
@xref{LOAD DATA, , @code{LOAD DATA}}.

In the resulting text file, only the following characters are escaped by
the @code{ESCAPED BY} character:

@itemize @bullet
@item The @code{ESCAPED BY} character
@item The first character in @code{FIELDS TERMINATED BY}
@item The first character in @code{LINES TERMINATED BY}
@end itemize

Additionally, @code{ASCII 0} is converted to @code{ESCAPED BY} followed by 0
(@code{ASCII 48}).

The reason for the above is that you @strong{must} escape any @code{FIELDS
TERMINATED BY}, @code{ESCAPED BY}, or @code{LINES TERMINATED BY}
characters to reliably be able to read the file back. @code{ASCII 0} is
escaped to make it easier to view with some pagers.

As the resulting file doesn't have to conform to the SQL syntax, nothing
else need be escaped.

Here follows an example of getting a file in the format used by many
old programs.

@example
SELECT a,b,a+b INTO OUTFILE "/tmp/result.text"
FIELDS TERMINATED BY ',' OPTIONALLY ENCLOSED BY '"'
LINES TERMINATED BY "\n"
FROM test_table;
@end example

@item
@findex DUMPFILE
If you use @code{INTO DUMPFILE} instead of @code{INTO OUTFILE}, MySQL
will only write one row into the file, without any column or line
terminations and without any escaping.  This is useful if you want to
store a blob in a file.

@item
Note that any file created by @code{INTO OUTFILE} and @code{INTO
DUMPFILE} is going to be readable for all users!  The reason is that the
MySQL server can't create a file that is owned by anyone else
than the user it's running as (you should never run @code{mysqld} as root),
the file has to be word readable so that you can retrieve the rows.

@item
If you are using @code{FOR UPDATE} on a table handler with page/row locks,
the examined rows will be write locked.
@end itemize


@menu
* JOIN::                        @code{JOIN} Syntax
* UNION::                       @code{UNION} Syntax
@end menu

@node JOIN, UNION, SELECT, SELECT
@subsubsection @code{JOIN} Syntax

@findex JOIN
@findex INNER JOIN
@findex CROSS JOIN
@findex LEFT JOIN
@findex LEFT OUTER JOIN
@findex NATURAL LEFT JOIN
@findex NATURAL LEFT OUTER JOIN
@findex RIGHT JOIN
@findex RIGHT OUTER JOIN
@findex NATURAL RIGHT JOIN
@findex NATURAL RIGHT OUTER JOIN
@findex STRAIGHT_JOIN

MySQL supports the following @code{JOIN} syntaxes for use in
@code{SELECT} statements:

@example
table_reference, table_reference
table_reference [CROSS] JOIN table_reference
table_reference INNER JOIN table_reference join_condition
table_reference STRAIGHT_JOIN table_reference
table_reference LEFT [OUTER] JOIN table_reference join_condition
table_reference LEFT [OUTER] JOIN table_reference
table_reference NATURAL [LEFT [OUTER]] JOIN table_reference
@{ oj table_reference LEFT OUTER JOIN table_reference ON conditional_expr @}
table_reference RIGHT [OUTER] JOIN table_reference join_condition
table_reference RIGHT [OUTER] JOIN table_reference
table_reference NATURAL [RIGHT [OUTER]] JOIN table_reference
@end example

Where @code{table_reference} is defined as:
@findex AS
@example
table_name [[AS] alias] [USE INDEX (key_list)] [IGNORE INDEX (key_list)]
@end example

and @code{join_condition} is defined as:

@example
ON conditional_expr |
USING (column_list)
@end example

You should never have any conditions in the @code{ON} part that are used to
restrict which rows you have in the result set.  If you want to restrict
which rows should be in the result, you have to do this in the @code{WHERE}
clause.

Note that in versions before Version 3.23.17, the @code{INNER JOIN} didn't
take a @code{join_condition}!

@cindex ODBC compatibility
@cindex compatibility, with ODBC
The last @code{LEFT OUTER JOIN} syntax shown above exists only for
compatibility with ODBC:

@itemize @bullet
@item
A table reference may be aliased using @code{tbl_name AS alias_name} or
@code{tbl_name alias_name}:

@example
mysql> SELECT t1.name, t2.salary FROM employee AS t1, info AS t2
    ->        WHERE t1.name = t2.name;
@end example

@item
The @code{ON} conditional is any conditional of the form that may be used in
a @code{WHERE} clause.

@item
If there is no matching record for the right table in the @code{ON} or
@code{USING} part in a @code{LEFT JOIN}, a row with all columns set to
@code{NULL} is used for the right table.  You can use this fact to find
records in a table that have no counterpart in another table:

@example
mysql> SELECT table1.* FROM table1
    ->        LEFT JOIN table2 ON table1.id=table2.id
    ->        WHERE table2.id IS NULL;
@end example

This example finds all rows in @code{table1} with an @code{id} value that is
not present in @code{table2} (that is, all rows in @code{table1} with no
corresponding row in @code{table2}).  This assumes that @code{table2.id} is
declared @code{NOT NULL}, of course. @xref{LEFT JOIN optimisation}.

@item
The @code{USING} @code{(column_list)} clause names a list of columns that must
exist in both tables.  A @code{USING} clause such as:

@example
A LEFT JOIN B USING (C1,C2,C3,...)
@end example

is defined to be semantically identical to an @code{ON} expression like
this:

@example
A.C1=B.C1 AND A.C2=B.C2 AND A.C3=B.C3,...
@end example

@item
The @code{NATURAL [LEFT] JOIN} of two tables is defined to be
semantically equivalent to an @code{INNER JOIN} or a @code{LEFT JOIN}
with a @code{USING} clause that names all columns that exist in both
tables.

@item
@cindex hints
@code{INNER JOIN} and @code{,} (comma) are semantically equivalent.
Both do a full join between the tables used. Normally, you specify
how the tables should be linked in the WHERE condition.

@item
@code{RIGHT JOIN} works analogously as @code{LEFT JOIN}.  To keep code
portable across databases, it's recommended to use @code{LEFT JOIN}
instead of @code{RIGHT JOIN}.

@item
@cindex hints
@code{STRAIGHT_JOIN} is identical to @code{JOIN}, except that the left table
is always read before the right table. This can be used for those (few)
cases where the join optimiser puts the tables in the wrong order.

@item
@cindex hints
@findex USE INDEX
@findex IGNORE INDEX
As of MySQL Version 3.23.12, you can give hints about which
index MySQL should use when retrieving information from a
table. This is useful if @code{EXPLAIN} shows that MySQL is
using the wrong index.  By specifying @code{USE INDEX (key_list)}, you
can tell MySQL to use only one of the specified indexes to
find rows in the table.  The alternative syntax @code{IGNORE INDEX
(key_list)} can be used to tell MySQL to not use some
particular index.
@findex USE KEY
@findex IGNORE KEY
@code{USE/IGNORE KEY} are synonyms for @code{USE/IGNORE INDEX}.
@end itemize

Some examples:

@example
mysql> SELECT * FROM table1,table2 WHERE table1.id=table2.id;
mysql> SELECT * FROM table1 LEFT JOIN table2 ON table1.id=table2.id;
mysql> SELECT * FROM table1 LEFT JOIN table2 USING (id);
mysql> SELECT * FROM table1 LEFT JOIN table2 ON table1.id=table2.id
    ->          LEFT JOIN table3 ON table2.id=table3.id;
mysql> SELECT * FROM table1 USE INDEX (key1,key2)
    ->          WHERE key1=1 AND key2=2 AND key3=3;
mysql> SELECT * FROM table1 IGNORE INDEX (key3)
    ->          WHERE key1=1 AND key2=2 AND key3=3;
@end example

@xref{LEFT JOIN optimisation, , @code{LEFT JOIN} optimisation}.


@node UNION,  , JOIN, SELECT
@subsubsection @code{UNION} Syntax

@findex UNION

@example
SELECT ...
UNION [ALL]
SELECT ...
  [UNION
   SELECT ...]
@end example

@code{UNION} is implemented in MySQL 4.0.0.

@code{UNION} is used to combine the result from many @code{SELECT}
statements into one result set.

The @code{SELECT} commands are normal select commands, but with the following
restrictions:

@itemize @bullet
@item
Only the last @code{SELECT} command can have @code{INTO OUTFILE}.
@end itemize

If you don't use the keyword @code{ALL} for the @code{UNION}, all
returned rows will be unique, as if you had done a @code{DISTINCT} for
the total result set.  If you specify @code{ALL}, then you will get all
matching rows from all the used @code{SELECT} statements.

If you want to use an @code{ORDER BY} for the total @code{UNION} result,
you should use parentheses:

@example
(SELECT a FROM table_name WHERE a=10 AND B=1 ORDER BY a LIMIT 10)
UNION
(SELECT a FROM table_name WHERE a=11 AND B=2 ORDER BY a LIMIT 10)
ORDER BY a;
@end example

@findex HANDLER
@node HANDLER, INSERT, SELECT, Data Manipulation
@subsection @code{HANDLER} Syntax

@example
HANDLER tbl_name OPEN [ AS alias ]
HANDLER tbl_name READ index_name @{ = | >= | <= | < @} (value1,value2,...)
    [ WHERE ... ] [LIMIT ... ]
HANDLER tbl_name READ index_name @{ FIRST | NEXT | PREV | LAST @}
    [ WHERE ... ] [LIMIT ... ]
HANDLER tbl_name READ @{ FIRST | NEXT @}
    [ WHERE ... ] [LIMIT ... ]
HANDLER tbl_name CLOSE
@end example

The @code{HANDLER} statement provides direct access to the @code{MyISAM} table
handler interface, bypassing the SQL optimiser. Thus, it is faster than
@code{SELECT}.

The first form of @code{HANDLER} statement opens a table, making
it accessible via subsequent @code{HANDLER ... READ} statements.
This table object is not shared by other threads and will not be closed
until the thread calls @code{HANDLER tbl_name CLOSE} or the thread dies.

@c FIX The sentence below needs to be rewritten. The reference to "the condition" is not clear.
@c FIX Also all mentions of the LIMIT clause need to be clarified in various sentences below.

The second form fetches one row (or more, specified by @code{LIMIT} clause)
where the index specified complies to the condition and @code{WHERE}
condition is met. If the index consists of several parts (spans over
several columns) the values are specified in comma-separated list,
providing values only for few first columns is possible.

The third form fetches one row (or more, specified by @code{LIMIT} clause)
from the table in index order, matching @code{WHERE} condition.

The fourth form (without index specification) fetches one row (or more, specified
by @code{LIMIT} clause) from the table in natural row order (as stored
in datafile) matching @code{WHERE} condition. It is faster than
@code{HANDLER tbl_name READ index_name} when a full table scan is desired.

@code{HANDLER ... CLOSE} closes a table that was opened with
@code{HANDLER ... OPEN}.

@code{HANDLER} is a somewhat low-level statement.  For example, it does
not provide consistency.  That is, @code{HANDLER ... OPEN} does @strong{NOT}
take a snapshot of the table, and does @strong{NOT} lock the table. This
means that after a @code{HANDLER ... OPEN} is issued, table data can be
modified (by this or any other thread) and these modifications may appear
only partially in @code{HANDLER ... NEXT} or @code{HANDLER ... PREV} scans.


@node INSERT, INSERT DELAYED, HANDLER, Data Manipulation
@subsection @code{INSERT} Syntax

@findex INSERT

@example
    INSERT [LOW_PRIORITY | DELAYED] [IGNORE]
        [INTO] tbl_name [(col_name,...)]
        VALUES (expression,...),(...),...
or  INSERT [LOW_PRIORITY | DELAYED] [IGNORE]
        [INTO] tbl_name [(col_name,...)]
        SELECT ...
or  INSERT [LOW_PRIORITY | DELAYED] [IGNORE]
        [INTO] tbl_name
        SET col_name=expression, col_name=expression, ...

@end example

@code{INSERT} inserts new rows into an existing table.  The @code{INSERT
...  VALUES} form of the statement inserts rows based on explicitly
specified values.  The @code{INSERT ... SELECT} form inserts rows
selected from another table or tables.  The @code{INSERT ... VALUES}
form with multiple value lists is supported in MySQL Version
3.22.5 or later.  The @code{col_name=expression} syntax is supported in
MySQL Version 3.22.10 or later.

@code{tbl_name} is the table into which rows should be inserted.  The column
name list or the @code{SET} clause indicates which columns the statement
specifies values for:

@itemize @bullet
@item
If you specify no column list for @code{INSERT ... VALUES} or @code{INSERT
... SELECT}, values for all columns must be provided in the
@code{VALUES()} list or by the @code{SELECT}.  If you don't know the order of
the columns in the table, use @code{DESCRIBE tbl_name} to find out.

@item
@cindex default values
Any column not explicitly given a value is set to its default value.  For
example, if you specify a column list that doesn't name all the columns in
the table, unnamed columns are set to their default values.  Default value
assignment is described in @ref{CREATE TABLE, , @code{CREATE TABLE}}.

MySQL always has a default value for all fields. This is something
that is imposed on MySQL to be able to work with both transactional
and not transactional tables.

Our view is that checking of fields content should be done in the
application and not in the database server.
@item
An @code{expression} may refer to any column that was set earlier in a value
list.  For example, you can say this:

@example
mysql> INSERT INTO tbl_name (col1,col2) VALUES(15,col1*2);
@end example

But not this:

@example
mysql> INSERT INTO tbl_name (col1,col2) VALUES(col2*2,15);
@end example

@item
If you specify the keyword @code{LOW_PRIORITY}, execution of the
@code{INSERT} is delayed until no other clients are reading from the
table.  In this case the client has to wait until the insert statement
is completed, which may take a long time if the table is in heavy
use. This is in contrast to @code{INSERT DELAYED}, which lets the client
continue at once.  @xref{INSERT DELAYED}.  Note that @code{LOW_PRIORITY}
should normally not be used with @code{MyISAM} tables as this disables
concurrent inserts. @xref{MyISAM}.

@item
If you specify the keyword @code{IGNORE} in an @code{INSERT} with many value
rows, any rows that duplicate an existing @code{PRIMARY} or @code{UNIQUE}
key in the table are ignored and are not inserted.  If you do not specify
@code{IGNORE}, the insert is aborted if there is any row that duplicates an
existing key value.  You can determine with the C API function
@code{mysql_info()} how many rows were inserted into the table.

@item
If MySQL was configured using the @code{DONT_USE_DEFAULT_FIELDS}
option, @code{INSERT} statements generate an error unless you explicitly
specify values for all columns that require a non-@code{NULL} value.
@xref{configure options,  , @code{configure} options}.

@item
You can find the value used for an @code{AUTO_INCREMENT} column
with the @code{mysql_insert_id} function.
@xref{mysql_insert_id, , @code{mysql_insert_id()}}.
@end itemize

@findex mysql_info()
If you use @code{INSERT ... SELECT} or an @code{INSERT ... VALUES}
statement with multiple value lists, you can use the C API function
@code{mysql_info()} to get information about the query.  The format of the
information string is shown here:

@example
Records: 100 Duplicates: 0 Warnings: 0
@end example

@code{Duplicates} indicates the number of rows that couldn't be inserted
because they would duplicate some existing unique index value.
@code{Warnings} indicates the number of attempts to insert column values that
were problematic in some way. Warnings can occur under any of the following
conditions:

@itemize @bullet
@item
Inserting @code{NULL} into a column that has been declared @code{NOT NULL}.
The column is set to its default value.

@item
Setting a numeric column to a value that lies outside the column's range.
The value is clipped to the appropriate endpoint of the range.

@item
Setting a numeric column to a value such as @code{'10.34 a'}.  The trailing
garbage is stripped and the remaining numeric part is inserted.  If the value
doesn't make sense as a number at all, the column is set to @code{0}.

@item
Inserting a string into a @code{CHAR}, @code{VARCHAR}, @code{TEXT}, or
@code{BLOB} column that exceeds the column's maximum length.  The value is
truncated to the column's maximum length.

@item
Inserting a value into a date or time column that is illegal for the column
type.  The column is set to the appropriate zero value for the type.
@end itemize

@findex REPLACE ... SELECT
@findex INSERT ... SELECT

@menu
* INSERT SELECT::               @code{INSERT ... SELECT} Syntax
@end menu


@node INSERT SELECT,  , INSERT, INSERT
@subsubsection @code{INSERT ... SELECT} Syntax

@example
INSERT [LOW_PRIORITY] [IGNORE] [INTO] tbl_name [(column list)] SELECT ...
@end example

With @code{INSERT ... SELECT} statement you can quickly insert many rows
into a table from one or many tables.

@example
INSERT INTO tblTemp2 (fldID) SELECT tblTemp1.fldOrder_ID FROM tblTemp1 WHERE
tblTemp1.fldOrder_ID > 100;
@end example

The following conditions hold for an @code{INSERT ... SELECT} statement:

@itemize @minus
@item
The target table of the @code{INSERT} statement cannot appear in the
@code{FROM} clause of the @code{SELECT} part of the query because it's
forbidden in ANSI SQL to @code{SELECT} from the same table into which you are
inserting.  (The problem is that the @code{SELECT} possibly would
find records that were inserted earlier during the same run.  When using
sub-select clauses, the situation could easily be very confusing!)

@item
@code{AUTO_INCREMENT} columns work as usual.

@item
You can use the C API function @code{mysql_info()} to get information about
the query. @xref{INSERT}.

@item
To ensure that the update log/binary log can be used to re-create the
original tables, MySQL will not allow concurrent inserts during
@code{INSERT ... SELECT}.
@end itemize

You can of course also use @code{REPLACE} instead of @code{INSERT} to
overwrite old rows.


@node INSERT DELAYED, UPDATE, INSERT, Data Manipulation
@subsection @code{INSERT DELAYED} Syntax

@findex INSERT DELAYED
@findex DELAYED

@cindex INSERT DELAYED

@example
INSERT DELAYED ...
@end example

The @code{DELAYED} option for the @code{INSERT} statement is a
MySQL-specific option that is very useful if you have clients
that can't wait for the @code{INSERT} to complete.  This is a common
problem when you use MySQL for logging and you also
periodically run @code{SELECT} and @code{UPDATE} statements that take a
long time to complete.  @code{DELAYED} was introduced in MySQL
Version 3.22.15.  It is a MySQL extension to ANSI SQL92.

@code{INSERT DELAYED} only works with @code{ISAM} and @code{MyISAM}
tables.  Note that as @code{MyISAM} tables supports concurrent
@code{SELECT} and @code{INSERT}, if there is no free blocks in the
middle of the datafile, you very seldom need to use @code{INSERT
DELAYED} with @code{MyISAM}. @xref{MyISAM}.

When you use @code{INSERT DELAYED}, the client will get an OK at once
and the row will be inserted when the table is not in use by any other thread.

Another major benefit of using @code{INSERT DELAYED} is that inserts
from many clients are bundled together and written in one block. This is much
faster than doing many separate inserts.

Note that currently the queued rows are only stored in memory until they are
inserted into the table.  This means that if you kill @code{mysqld}
the hard way (@code{kill -9}) or if @code{mysqld} dies unexpectedly, any
queued rows that weren't written to disk are lost!

The following describes in detail what happens when you use the
@code{DELAYED} option to @code{INSERT} or @code{REPLACE}.  In this
description, the ``thread'' is the thread that received an @code{INSERT
DELAYED} command and ``handler'' is the thread that handles all
@code{INSERT DELAYED} statements for a particular table.

@itemize @bullet
@item
When a thread executes a @code{DELAYED} statement for a table, a handler
thread is created to process all @code{DELAYED} statements for the table, if
no such handler already exists.

@item
The thread checks whether the handler has acquired a @code{DELAYED}
lock already; if not, it tells the handler thread to do so.  The
@code{DELAYED} lock can be obtained even if other threads have a @code{READ}
or @code{WRITE} lock on the table.  However, the handler will wait for all
@code{ALTER TABLE} locks or @code{FLUSH TABLES} to ensure that the table
structure is up to date.

@item
The thread executes the @code{INSERT} statement, but instead of writing
the row to the table, it puts a copy of the final row into a queue that
is managed by the handler thread. Any syntax errors are noticed by the
thread and reported to the client program.

@item
The client can't report the number of duplicates or the @code{AUTO_INCREMENT}
value for the resulting row; it can't obtain them from the server, because
the @code{INSERT} returns before the insert operation has been completed.  If
you use the C API, the @code{mysql_info()} function doesn't return anything
meaningful, for the same reason.

@item
The update log is updated by the handler thread when the row is inserted into
the table.  In case of multiple-row inserts, the update log is updated when
the first row is inserted.

@item
After every @code{delayed_insert_limit} rows are written, the handler checks
whether any @code{SELECT} statements are still pending.  If so, it
allows these to execute before continuing.

@cindex delayed_insert_limit
@cindex timeout
@item
When the handler has no more rows in its queue, the table is unlocked.  If no
new @code{INSERT DELAYED} commands are received within
@code{delayed_insert_timeout} seconds, the handler terminates.

@item
If more than @code{delayed_queue_size} rows are pending already in a
specific handler queue, the thread requesting @code{INSERT DELAYED}
waits until there is room in the queue.  This is done to ensure that
the @code{mysqld} server doesn't use all memory for the delayed memory
queue.

@item
The handler thread will show up in the MySQL process list
with @code{delayed_insert} in the @code{Command} column.  It will
be killed if you execute a @code{FLUSH TABLES} command or kill it with
@code{KILL thread_id}. However, it will first store all queued rows into the
table before exiting. During this time it will not accept any new
@code{INSERT} commands from another thread. If you execute an @code{INSERT
DELAYED} command after this, a new handler thread will be created.

Note that the above means that @code{INSERT DELAYED} commands have higher
priority than normal @code{INSERT} commands if there is an @code{INSERT
DELAYED} handler already running!  Other update commands will have to wait
until the @code{INSERT DELAYED} queue is empty, someone kills the handler
thread (with @code{KILL thread_id}), or someone executes @code{FLUSH TABLES}.

@item
The following status variables provide information about @code{INSERT
DELAYED} commands:

@multitable @columnfractions .30 .55
@item @strong{Variable} @tab @strong{Meaning}
@item @code{Delayed_insert_threads} @tab Number of handler threads
@item @code{Delayed_writes} @tab Number of rows written with @code{INSERT DELAYED}
@item @code{Not_flushed_delayed_rows} @tab Number of rows waiting to be written
@end multitable

You can view these variables by issuing a @code{SHOW STATUS} statement or
by executing a @code{mysqladmin extended-status} command.
@end itemize

Note that @code{INSERT DELAYED} is slower than a normal INSERT if the
table is not in use.  There is also the additional overhead for the
server to handle a separate thread for each table on which you use
@code{INSERT DELAYED}.  This means that you should only use @code{INSERT
DELAYED} when you are really sure you need it!


@node UPDATE, DELETE, INSERT DELAYED, Data Manipulation
@subsection @code{UPDATE} Syntax

@findex UPDATE

@example
UPDATE [LOW_PRIORITY] [IGNORE] tbl_name
    SET col_name1=expr1 [, col_name2=expr2, ...]
    [WHERE where_definition]
    [LIMIT #]
@end example

@code{UPDATE} updates columns in existing table rows with new values.
The @code{SET} clause indicates which columns to modify and the values
they should be given.  The @code{WHERE} clause, if given, specifies
which rows should be updated.  Otherwise, all rows are updated. If the
@code{ORDER BY} clause is specified, the rows will be updated in the
order that is specified.

If you specify the keyword @code{LOW_PRIORITY}, execution of the
@code{UPDATE} is delayed until no other clients are reading from the table.

If you specify the keyword @code{IGNORE}, the update statement will not
abort even if we get duplicate key errors during the update.  Rows that
would cause conflicts will not be updated.

If you access a column from @code{tbl_name} in an expression,
@code{UPDATE} uses the current value of the column.  For example, the
following statement sets the @code{age} column to one more than its
current value:

@example
mysql> UPDATE persondata SET age=age+1;
@end example

@code{UPDATE} assignments are evaluated from left to right.  For example, the
following statement doubles the @code{age} column, then increments it:

@example
mysql> UPDATE persondata SET age=age*2, age=age+1;
@end example

If you set a column to the value it currently has, MySQL notices
this and doesn't update it.

@findex mysql_info()
@code{UPDATE} returns the number of rows that were actually changed.
In MySQL Version 3.22 or later, the C API function @code{mysql_info()}
returns the number of rows that were matched and updated and the number of
warnings that occurred during the @code{UPDATE}.

In MySQL Version 3.23, you can use @code{LIMIT #} to ensure that
only a given number of rows are changed.


@node DELETE, TRUNCATE, UPDATE, Data Manipulation
@subsection @code{DELETE} Syntax

@findex DELETE

@example
DELETE [LOW_PRIORITY | QUICK] FROM table_name
       [WHERE where_definition]
       [ORDER BY ...]
       [LIMIT rows]

or

DELETE [LOW_PRIORITY | QUICK] table_name[.*] [,table_name[.*] ...]
       FROM table-references
       [WHERE where_definition]

or

DELETE [LOW_PRIORITY | QUICK]
       FROM table_name[.*], [table_name[.*] ...]
       USING table-references
       [WHERE where_definition]
@end example

@code{DELETE} deletes rows from @code{table_name} that satisfy the condition
given by @code{where_definition}, and returns the number of records deleted.

If you issue a @code{DELETE} with no @code{WHERE} clause, all rows are
deleted.  If you do this in @code{AUTOCOMMIT} mode, this works as
@code{TRUNCATE}. @xref{TRUNCATE, , @code{TRUNCATE}}.  In MySQL 3.23,
@code{DELETE} without a @code{WHERE} clause will return zero as the number
of affected records.

If you really want to know how many records are deleted when you are deleting
all rows, and are willing to suffer a speed penalty, you can use a
@code{DELETE} statement of this form:

@example
mysql> DELETE FROM table_name WHERE 1>0;
@end example

Note that this is much slower than @code{DELETE FROM table_name} with no
@code{WHERE} clause, because it deletes rows one at a time.

If you specify the keyword @code{LOW_PRIORITY}, execution of the
@code{DELETE} is delayed until no other clients are reading from the table.

If you specify the word @code{QUICK} then the table handler will not
merge index leaves during delete, which may speed up certain kind of
deletes.

In @code{MyISAM} tables, deleted records are maintained in a linked list and
subsequent @code{INSERT} operations reuse old record positions. To
reclaim unused space and reduce file-sizes, use the @code{OPTIMIZE
TABLE} statement or the @code{myisamchk} utility to reorganise tables.
@code{OPTIMIZE TABLE} is easier, but @code{myisamchk} is faster.  See
@ref{OPTIMIZE TABLE, , @code{OPTIMIZE TABLE}} and @ref{Optimisation}.

The first multi-table delete format is supported starting from MySQL 4.0.0.
The second multi-table delete format is supported starting from MySQL 4.0.2.

The idea is that only matching rows from the tables listed
@strong{before} the @code{FROM} or before the @code{USING} clause are
deleted. The effect is that you can delete rows from many tables at the
same time and also have additional tables that are used for searching.

The @code{.*} after the table names is there just to be compatible with
@code{Access}:

@example
DELETE t1,t2 FROM t1,t2,t3 WHERE t1.id=t2.id AND t2.id=t3.id

or

DELETE FROM t1,t2 USING t1,t2,t3 WHERE t1.id=t2.id AND t2.id=t3.id
@end example

In the above case we delete matching rows just from tables @code{t1} and
@code{t2}.

@code{ORDER BY} and using multiple tables in the @code{DELETE} statement
is supported in MySQL 4.0.

If an @code{ORDER BY} clause is used, the rows will be deleted in that order.
This is really only useful in conjunction with @code{LIMIT}.  For example:

@example
DELETE FROM somelog
WHERE user = 'jcole'
ORDER BY timestamp
LIMIT 1
@end example

This will delete the oldest entry (by @code{timestamp}) where the row matches
the @code{WHERE} clause.

The MySQL-specific @code{LIMIT rows} option to @code{DELETE} tells
the server the maximum number of rows to be deleted before control is
returned to the client.  This can be used to ensure that a specific
@code{DELETE} command doesn't take too much time.  You can simply repeat
the @code{DELETE} command until the number of affected rows is less than
the @code{LIMIT} value.


@node TRUNCATE, REPLACE, DELETE, Data Manipulation
@subsection @code{TRUNCATE} Syntax

@findex TRUNCATE

@example
TRUNCATE TABLE table_name
@end example

In 3.23 @code{TRUNCATE TABLE} is mapped to
@code{COMMIT ; DELETE FROM table_name}. @xref{DELETE}.

@code{TRUNCATE TABLE} differs from @code{DELETE FROM ...}
in the following ways:

@itemize @bullet
@item
Truncate operations drop and re-create the table, which is much faster
than deleting rows one by one.
@item
Not transaction-safe; you will get an error if you have an active
transaction or an active table lock.
@item
Doesn't return the number of deleted rows.
@item
As long as the table definition file @file{table_name.frm} is
valid, the table can be re-created this way, even if the data or index
files have become corrupted.
@end itemize

@code{TRUNCATE} is an Oracle SQL extension.

@node REPLACE, LOAD DATA, TRUNCATE, Data Manipulation
@subsection @code{REPLACE} Syntax

@findex REPLACE

@example
    REPLACE [LOW_PRIORITY | DELAYED]
        [INTO] tbl_name [(col_name,...)]
        VALUES (expression,...),(...),...
or  REPLACE [LOW_PRIORITY | DELAYED]
        [INTO] tbl_name [(col_name,...)]
        SELECT ...
or  REPLACE [LOW_PRIORITY | DELAYED]
        [INTO] tbl_name
        SET col_name=expression, col_name=expression,...
@end example

@code{REPLACE} works exactly like @code{INSERT}, except that if an old
record in the table has the same value as a new record on a unique index,
the old record is deleted before the new record is inserted.
@xref{INSERT, , @code{INSERT}}.

In other words, you can't access the values of the old row from a
@code{REPLACE} statement.  In some old MySQL versions it appeared that
you could do this, but that was a bug that has been corrected.

When you use a @code{REPLACE} command, @code{mysql_affected_rows()}
will return 2 if the new row replaced an old row.  This is because
one row was inserted and then the duplicate was deleted.

This fact makes it easy to determine whether @code{REPLACE} added
or replaced a row: check whether the affected-rows value is 1 (added)
or 2 (replaced).


@node LOAD DATA, DO, REPLACE, Data Manipulation
@subsection @code{LOAD DATA INFILE} Syntax

@findex LOAD DATA INFILE

@example
LOAD DATA [LOW_PRIORITY | CONCURRENT] [LOCAL] INFILE 'file_name.txt'
    [REPLACE | IGNORE]
    INTO TABLE tbl_name
    [FIELDS
        [TERMINATED BY '\t']
        [[OPTIONALLY] ENCLOSED BY '']
        [ESCAPED BY '\\' ]
    ]
    [LINES TERMINATED BY '\n']
    [IGNORE number LINES]
    [(col_name,...)]
@end example

The @code{LOAD DATA INFILE} statement reads rows from a text file into a
table at a very high speed.  If the @code{LOCAL} keyword is specified, the
file is read from the client host.  If @code{LOCAL} is not specified, the
file must be located on the server.  (@code{LOCAL} is available in
MySQL Version 3.22.6 or later.)

For security reasons, when reading text files located on the server, the
files must either reside in the database directory or be readable by all.
Also, to use @code{LOAD DATA INFILE} on server files, you must have the
@strong{file} privilege on the server host.
@xref{Privileges provided}.

In MySQL 3.23.49 and MySQL 4.0.2 @code{LOCAL} will only work if you have
not started @code{mysqld} with @code{--local-infile=0} or if you
have not enabled your client to support @code{LOCAL}. @xref{LOAD DATA LOCAL}.

If you specify the keyword @code{LOW_PRIORITY}, execution of the
@code{LOAD DATA} statement is delayed until no other clients are reading
from the table.

If you specify the keyword @code{CONCURRENT} with a @code{MyISAM} table,
then other threads can retrieve data from the table while @code{LOAD
DATA} is executing. Using this option will of course affect the
performance of @code{LOAD DATA} a bit even if no other thread is using
the table at the same time.

Using @code{LOCAL} will be a bit slower than letting the server access the
files directly, because the contents of the file must travel from the client
host to the server host.  On the other hand, you do not need the
@strong{file} privilege to load local files.

@c old version
If you are using MySQL before Version 3.23.24 you can't read from a
FIFO with @code{LOAD DATA INFILE}.  If you need to read from a FIFO (for
example the output from gunzip), use @code{LOAD DATA LOCAL INFILE}
instead.

@cindex @code{mysqlimport}
You can also load datafiles by using the @code{mysqlimport} utility; it
operates by sending a @code{LOAD DATA INFILE} command to the server.  The
@code{--local} option causes @code{mysqlimport} to read datafiles from the
client host.  You can specify the @code{--compress} option to get better
performance over slow networks if the client and server support the
compressed protocol.

When locating files on the server host, the server uses the following rules:

@itemize @bullet
@item
If an absolute pathname is given, the server uses the pathname as is.

@item
If a relative pathname with one or more leading components is given,
the server searches for the file relative to the server's data directory.

@item
If a filename with no leading components is given, the server looks for
the file in the database directory of the current database.
@end itemize

Note that these rules mean a file given as @file{./myfile.txt} is read from
the server's data directory, whereas a file given as @file{myfile.txt} is
read from the database directory of the current database.  For example,
the following @code{LOAD DATA} statement reads the file @file{data.txt}
from the database directory for @code{db1} because @code{db1} is the current
database, even though the statement explicitly loads the file into a
table in the @code{db2} database:

@example
mysql> USE db1;
mysql> LOAD DATA INFILE "data.txt" INTO TABLE db2.my_table;
@end example

The @code{REPLACE} and @code{IGNORE} keywords control handling of input
records that duplicate existing records on unique key values.  If you specify
@code{REPLACE}, new rows replace existing rows that have the same unique key
value. If you specify @code{IGNORE}, input rows that duplicate an existing
row on a unique key value are skipped.  If you don't specify either option, an
error occurs when a duplicate key value is found, and the rest of the text
file is ignored.

If you load data from a local file using the @code{LOCAL} keyword, the server
has no way to stop transmission of the file in the middle of the operation,
so the default bahavior is the same as if @code{IGNORE} is specified.

If you use @code{LOAD DATA INFILE} on an empty @code{MyISAM} table,
all non-unique indexes are created in a separate batch (like in @code{REPAIR}).
This normally makes @code{LOAD DATA INFILE} much faster when you have many
indexes.

@code{LOAD DATA INFILE} is the complement of @code{SELECT ... INTO OUTFILE}.
@xref{SELECT, , @code{SELECT}}.
To write data from a database to a file, use @code{SELECT ... INTO OUTFILE}.
To read the file back into the database, use @code{LOAD DATA INFILE}.
The syntax of the @code{FIELDS} and @code{LINES} clauses is the same for
both commands.  Both clauses are optional, but @code{FIELDS}
must precede @code{LINES} if both are specified.

If you specify a @code{FIELDS} clause,
each of its subclauses (@code{TERMINATED BY}, @code{[OPTIONALLY] ENCLOSED
BY}, and @code{ESCAPED BY}) is also optional, except that you must
specify at least one of them.

If you don't specify a @code{FIELDS} clause, the defaults are the
same as if you had written this:

@example
FIELDS TERMINATED BY '\t' ENCLOSED BY '' ESCAPED BY '\\'
@end example

If you don't specify a @code{LINES} clause, the default
is the same as if you had written this:

@example
LINES TERMINATED BY '\n'
@end example

In other words, the defaults cause @code{LOAD DATA INFILE} to act as follows
when reading input:

@itemize @bullet
@item
Look for line boundaries at newlines.

@item
Break lines into fields at tabs.

@item
Do not expect fields to be enclosed within any quoting characters.

@item
Interpret occurrences of tab, newline, or @samp{\} preceded by
@samp{\} as literal characters that are part of field values.
@end itemize

Conversely, the defaults cause @code{SELECT ... INTO OUTFILE} to act as
follows when writing output:

@itemize @bullet
@item
Write tabs between fields.

@item
Do not enclose fields within any quoting characters.

@item
Use @samp{\} to escape instances of tab, newline or @samp{\} that occur
within field values.

@item
Write newlines at the ends of lines.
@end itemize

Note that to write @code{FIELDS ESCAPED BY '\\'}, you must specify two
backslashes for the value to be read as a single backslash.

The @code{IGNORE number LINES} option can be used to ignore a header of
column names at the start of the file:

@example
mysql> LOAD DATA INFILE "/tmp/file_name" INTO TABLE test IGNORE 1 LINES;
@end example

When you use @code{SELECT ... INTO OUTFILE} in tandem with @code{LOAD
DATA INFILE} to write data from a database into a file and then read
the file back into the database later, the field and line handling
options for both commands must match.  Otherwise, @code{LOAD DATA
INFILE} will not interpret the contents of the file properly.  Suppose
you use @code{SELECT ... INTO OUTFILE} to write a file with
fields delimited by commas:

@example
mysql> SELECT * INTO OUTFILE 'data.txt'
    ->          FIELDS TERMINATED BY ','
    ->          FROM ...;
@end example

To read the comma-delimited file back in, the correct statement would be:

@example
mysql> LOAD DATA INFILE 'data.txt' INTO TABLE table2
    ->           FIELDS TERMINATED BY ',';
@end example

If instead you tried to read in the file with the statement shown here, it
wouldn't work because it instructs @code{LOAD DATA INFILE} to look for
tabs between fields:

@example
mysql> LOAD DATA INFILE 'data.txt' INTO TABLE table2
    ->           FIELDS TERMINATED BY '\t';
@end example

The likely result is that each input line would be interpreted as
a single field.

@code{LOAD DATA INFILE} can be used to read files obtained from
external sources, too. For example, a file in dBASE format will have
fields separated by commas and enclosed in double quotes.  If lines in
the file are terminated by newlines, the command shown here
illustrates the field and line handling options you would use to load
the file:

@example
mysql> LOAD DATA INFILE 'data.txt' INTO TABLE tbl_name
    ->           FIELDS TERMINATED BY ',' ENCLOSED BY '"'
    ->           LINES TERMINATED BY '\n';
@end example

Any of the field or line handling options may specify an empty string
(@code{''}).  If not empty, the @code{FIELDS [OPTIONALLY] ENCLOSED BY}
and @code{FIELDS ESCAPED BY} values must be a single character.  The
@code{FIELDS TERMINATED BY} and @code{LINES TERMINATED BY} values may
be more than one character.  For example, to write lines that are
terminated by carriage return-linefeed pairs, or to read a file
containing such lines, specify a @code{LINES TERMINATED BY '\r\n'}
clause.

For example, to read a file of jokes, that are separated with a line
of @code{%%}, into a SQL table you can do:

@example
CREATE TABLE jokes (a INT NOT NULL AUTO_INCREMENT PRIMARY KEY, joke TEXT
NOT NULL);
LOAD DATA INFILE "/tmp/jokes.txt" INTO TABLE jokes FIELDS TERMINATED BY ""
LINES TERMINATED BY "\n%%\n" (joke);
@end example

@code{FIELDS [OPTIONALLY] ENCLOSED BY} controls quoting of fields.  For
output (@code{SELECT ... INTO OUTFILE}), if you omit the word
@code{OPTIONALLY}, all fields are enclosed by the @code{ENCLOSED BY}
character.  An example of such output (using a comma as the field
delimiter) is shown here:

@example
"1","a string","100.20"
"2","a string containing a , comma","102.20"
"3","a string containing a \" quote","102.20"
"4","a string containing a \", quote and comma","102.20"
@end example

If you specify @code{OPTIONALLY}, the @code{ENCLOSED BY} character is
used only to enclose @code{CHAR} and @code{VARCHAR} fields:

@example
1,"a string",100.20
2,"a string containing a , comma",102.20
3,"a string containing a \" quote",102.20
4,"a string containing a \", quote and comma",102.20
@end example

Note that occurrences of the @code{ENCLOSED BY} character within a
field value are escaped by prefixing them with the @code{ESCAPED BY}
character.  Also note that if you specify an empty @code{ESCAPED BY}
value, it is possible to generate output that cannot be read properly by
@code{LOAD DATA INFILE}. For example, the output just shown above would
appear as shown here if the escape character is empty.  Observe that the
second field in the fourth line contains a comma following the quote, which
(erroneously) appears to terminate the field:

@example
1,"a string",100.20
2,"a string containing a , comma",102.20
3,"a string containing a " quote",102.20
4,"a string containing a ", quote and comma",102.20
@end example

For input, the @code{ENCLOSED BY} character, if present, is stripped from the
ends of field values.  (This is true whether @code{OPTIONALLY} is
specified; @code{OPTIONALLY} has no effect on input interpretation.)
Occurrences of the @code{ENCLOSED BY} character preceded by the
@code{ESCAPED BY} character are interpreted as part of the current field
value.  In addition, duplicated @code{ENCLOSED BY} characters occurring
within fields are interpreted as single @code{ENCLOSED BY} characters if the
field itself starts with that character.  For example, if @code{ENCLOSED BY
'"'} is specified, quotes are handled as shown here:

@example
"The ""BIG"" boss"  -> The "BIG" boss
The "BIG" boss      -> The "BIG" boss
The ""BIG"" boss    -> The ""BIG"" boss
@end example

@code{FIELDS ESCAPED BY} controls how to write or read special characters.
If the @code{FIELDS ESCAPED BY} character is not empty, it is used to prefix
the following characters on output:
@itemize @bullet
@item
The @code{FIELDS ESCAPED BY} character
@item
The @code{FIELDS [OPTIONALLY] ENCLOSED BY} character
@item
The first character of the @code{FIELDS TERMINATED BY} and
@code{LINES TERMINATED BY} values
@item
ASCII @code{0} (what is actually written following the escape character is
ASCII @code{'0'}, not a zero-valued byte)
@end itemize

If the @code{FIELDS ESCAPED BY} character is empty, no characters are escaped.
It is probably not a good idea to specify an empty escape character,
particularly if field values in your data contain any of the characters in
the list just given.

For input, if the @code{FIELDS ESCAPED BY} character is not empty, occurrences
of that character are stripped and the following character is taken literally
as part of a field value.  The exceptions are an escaped @samp{0} or
@samp{N} (for example, @code{\0} or @code{\N} if the escape character is
@samp{\}).  These sequences are interpreted as ASCII @code{0} (a zero-valued
byte) and @code{NULL}.  See below for the rules on @code{NULL} handling.

For more information about @samp{\}-escape syntax,
see @ref{Literals}.

In certain cases, field and line handling options interact:

@itemize @bullet
@item
If @code{LINES TERMINATED BY} is an empty string and @code{FIELDS
TERMINATED BY} is non-empty, lines are also terminated with
@code{FIELDS TERMINATED BY}.
@item
If the @code{FIELDS TERMINATED BY} and @code{FIELDS ENCLOSED BY} values are
both empty (@code{''}), a fixed-row (non-delimited) format is used.  With
fixed-row format, no delimiters are used between fields. Instead, column
values are written and read using the ``display'' widths of the columns.  For
example, if a column is declared as @code{INT(7)}, values for the column are
written using 7-character fields.  On input, values for the column are
obtained by reading 7 characters.  Fixed-row format also affects handling of
@code{NULL} values; see below.  Note that fixed-size format will not work
if you are using a multi-byte character set.
@end itemize

Handling of @code{NULL} values varies, depending on the @code{FIELDS} and
@code{LINES} options you use:

@itemize @bullet
@item
For the default @code{FIELDS} and @code{LINES} values,
@code{NULL} is written as @code{\N} for output and @code{\N} is read
as @code{NULL} for input (assuming the @code{ESCAPED BY} character
is @samp{\}).

@item
If @code{FIELDS ENCLOSED BY} is not empty, a field containing the literal
word @code{NULL} as its value is read as a @code{NULL} value (this differs
from the word @code{NULL} enclosed within @code{FIELDS ENCLOSED BY}
characters, which is read as the string @code{'NULL'}).

@item
If @code{FIELDS ESCAPED BY} is empty, @code{NULL} is written as the word
@code{NULL}.

@item
With fixed-row format (which happens when @code{FIELDS TERMINATED BY} and
@code{FIELDS ENCLOSED BY} are both empty), @code{NULL} is written as an empty
string.  Note that this causes both @code{NULL} values and empty strings in
the table to be indistinguishable when written to the file because they are
both written as empty strings.  If you need to be able to tell the two apart
when reading the file back in, you should not use fixed-row format.
@end itemize

Some cases are not supported by @code{LOAD DATA INFILE}:
@itemize @bullet

@item
Fixed-size rows (@code{FIELDS TERMINATED BY} and @code{FIELDS ENCLOSED
BY} both empty) and @code{BLOB} or @code{TEXT} columns.

@item
If you specify one separator that is the same as or a prefix of another,
@code{LOAD DATA INFILE} won't be able to interpret the input properly.
For example, the following @code{FIELDS} clause would cause problems:

@example
FIELDS TERMINATED BY '"' ENCLOSED BY '"'
@end example

@item
If @code{FIELDS ESCAPED BY} is empty, a field value that contains an occurrence
of @code{FIELDS ENCLOSED BY} or @code{LINES TERMINATED BY}
followed by the @code{FIELDS TERMINATED BY} value will cause @code{LOAD
DATA INFILE} to stop reading a field or line too early.
This happens because @code{LOAD DATA INFILE} cannot properly determine
where the field or line value ends.
@end itemize

The following example loads all columns of the @code{persondata} table:

@example
mysql> LOAD DATA INFILE 'persondata.txt' INTO TABLE persondata;
@end example

No field list is specified, so @code{LOAD DATA INFILE} expects input rows
to contain a field for each table column.  The default @code{FIELDS} and
@code{LINES} values are used.

If you wish to load only some of a table's columns, specify a field list:

@example
mysql> LOAD DATA INFILE 'persondata.txt'
    ->           INTO TABLE persondata (col1,col2,...);
@end example

You must also specify a field list if the order of the fields in the input
file differs from the order of the columns in the table.  Otherwise,
MySQL cannot tell how to match up input fields with table columns.

If a row has too few fields, the columns for which no input field is present
are set to default values.  Default value assignment is described in
@ref{CREATE TABLE, , @code{CREATE TABLE}}.

An empty field value is interpreted differently than if the field value
is missing:

@itemize @bullet
@item
For string types, the column is set to the empty string.

@item
For numeric types, the column is set to @code{0}.

@item
For date and time types, the column is set to the appropriate ``zero''
value for the type.
@xref{Date and time types}.
@end itemize

Note that these are the same values that result if you assign an empty
string explicitly to a string, numeric, or date or time type explicitly
in an @code{INSERT} or @code{UPDATE} statement.

@code{TIMESTAMP} columns are only set to the current date and time if there
is a @code{NULL} value for the column, or (for the first @code{TIMESTAMP}
column only) if the @code{TIMESTAMP} column is left out from the field list
when a field list is specified.

If an input row has too many fields, the extra fields are ignored and
the number of warnings is incremented.

@code{LOAD DATA INFILE} regards all input as strings, so you can't use
numeric values for @code{ENUM} or @code{SET} columns the way you can with
@code{INSERT} statements.  All @code{ENUM} and @code{SET} values must be
specified as strings!

@findex mysql_info()
If you are using the C API, you can get information about the query by
calling the API function @code{mysql_info()} when the @code{LOAD DATA INFILE}
query finishes.  The format of the information string is shown here:

@example
Records: 1  Deleted: 0  Skipped: 0  Warnings: 0
@end example

Warnings occur under the same circumstances as when values are inserted
via the @code{INSERT} statement (@pxref{INSERT, , @code{INSERT}}), except
that @code{LOAD DATA INFILE} also generates warnings when there are too few
or too many fields in the input row.  The warnings are not stored anywhere;
the number of warnings can only be used as an indication if everything went
well.  If you get warnings and want to know exactly why you got them, one way
to do this is to use @code{SELECT ... INTO OUTFILE} into another file and
compare this to your original input file.

If you need @code{LOAD DATA} to read from a pipe, you can use the
following trick:

@example
mkfifo /mysql/db/x/x
chmod 666 /mysql/db/x/x
cat < /dev/tcp/10.1.1.12/4711 > /nt/mysql/db/x/x
mysql -e "LOAD DATA INFILE 'x' INTO TABLE x" x
@end example

If you are using a version of MySQL older than 3.23.25
you can only do the above with @code{LOAD DATA LOCAL INFILE}.

For more information about the efficiency of @code{INSERT} versus
@code{LOAD DATA INFILE} and speeding up @code{LOAD DATA INFILE},
@xref{Insert speed}.


@node DO,  , LOAD DATA, Data Manipulation
@subsection @code{DO} Syntax

@findex DO

@example
DO expression, [expression, ...]
@end example

Execute the expression but don't return any results.  This is a
shorthand of @code{SELECT expression, expression}, but has the advantage
that it's slightly faster when you don't care about the result.

This is mainly useful with functions that has side effects, like
@code{RELEASE_LOCK}.


@node Data Definition, Basic User Commands, Data Manipulation, Reference
@section Data Definition: @code{CREATE}, @code{DROP}, @code{ALTER}

@menu
* CREATE DATABASE::             @code{CREATE DATABASE} Syntax
* DROP DATABASE::               @code{DROP DATABASE} Syntax
* CREATE TABLE::                @code{CREATE TABLE} Syntax
* ALTER TABLE::                 @code{ALTER TABLE} Syntax
* RENAME TABLE::                @code{RENAME TABLE} Syntax
* DROP TABLE::                  @code{DROP TABLE} Syntax
* CREATE INDEX::                @code{CREATE INDEX} Syntax
* DROP INDEX::                  @code{DROP INDEX} Syntax
@end menu


@node CREATE DATABASE, DROP DATABASE, Data Definition, Data Definition
@subsection @code{CREATE DATABASE} Syntax

@findex CREATE DATABASE

@example
CREATE DATABASE [IF NOT EXISTS] db_name
@end example

@code{CREATE DATABASE} creates a database with the given name.  Rules for
allowable database names are given in @ref{Legal names}.  An error occurs if
the database already exists and you didn't specify @code{IF NOT EXISTS}.

Databases in MySQL are implemented as directories containing files
that correspond to tables in the database.  Because there are no tables in a
database when it is initially created, the @code{CREATE DATABASE} statement
only creates a directory under the MySQL data directory.

@cindex @code{mysqladmin}
You can also create databases with @code{mysqladmin}.
@xref{Client-Side Scripts}.


@node DROP DATABASE, CREATE TABLE, CREATE DATABASE, Data Definition
@subsection @code{DROP DATABASE} Syntax

@findex DROP DATABASE

@example
DROP DATABASE [IF EXISTS] db_name
@end example

@code{DROP DATABASE} drops all tables in the database and deletes the
database.  If you do a @code{DROP DATABASE} on a symbolic linked
database, both the link and the original database is deleted. @strong{Be
VERY careful with this command!}

@code{DROP DATABASE} returns the number of files that were removed from
the database directory.  Normally, this is three times the number of
tables, because normally each table corresponds to a @file{.MYD} file, a
@file{.MYI} file, and a @file{.frm} file.

The @code{DROP DATABASE} command removes from the given database
directory all files with the following extensions:

@multitable @columnfractions .10 .10 .10 .10
@item @strong{Ext} @tab @strong{Ext} @tab @strong{Ext} @tab @strong{Ext}
@item .BAK  @tab .DAT  @tab .HSH  @tab .ISD
@item .ISM  @tab .ISM  @tab .MRG  @tab .MYD
@item .MYI  @tab .db   @tab .frm  @tab
@end multitable

All subdirectories that consists of 2 digits (@code{RAID} directories)
are also removed.

In MySQL Version 3.22 or later, you can use the keywords
@code{IF EXISTS} to prevent an error from occurring if the database doesn't
exist.

@cindex @code{mysqladmin}
You can also drop databases with @code{mysqladmin}. @xref{Client-Side Scripts}.

@node CREATE TABLE, ALTER TABLE, DROP DATABASE, Data Definition
@subsection @code{CREATE TABLE} Syntax

@findex CREATE TABLE

@menu
* Silent column changes::       Silent column changes
@end menu

@example
CREATE [TEMPORARY] TABLE [IF NOT EXISTS] tbl_name [(create_definition,...)]
[table_options] [select_statement]

create_definition:
  col_name type [NOT NULL | NULL] [DEFAULT default_value] [AUTO_INCREMENT]
            [PRIMARY KEY] [reference_definition]
  or    PRIMARY KEY (index_col_name,...)
  or    KEY [index_name] (index_col_name,...)
  or    INDEX [index_name] (index_col_name,...)
  or    UNIQUE [INDEX] [index_name] (index_col_name,...)
  or    FULLTEXT [INDEX] [index_name] (index_col_name,...)
  or    [CONSTRAINT symbol] FOREIGN KEY [index_name] (index_col_name,...)
            [reference_definition]
  or    CHECK (expr)

type:
        TINYINT[(length)] [UNSIGNED] [ZEROFILL]
  or    SMALLINT[(length)] [UNSIGNED] [ZEROFILL]
  or    MEDIUMINT[(length)] [UNSIGNED] [ZEROFILL]
  or    INT[(length)] [UNSIGNED] [ZEROFILL]
  or    INTEGER[(length)] [UNSIGNED] [ZEROFILL]
  or    BIGINT[(length)] [UNSIGNED] [ZEROFILL]
  or    REAL[(length,decimals)] [UNSIGNED] [ZEROFILL]
  or    DOUBLE[(length,decimals)] [UNSIGNED] [ZEROFILL]
  or    FLOAT[(length,decimals)] [UNSIGNED] [ZEROFILL]
  or    DECIMAL(length,decimals) [UNSIGNED] [ZEROFILL]
  or    NUMERIC(length,decimals) [UNSIGNED] [ZEROFILL]
  or    CHAR(length) [BINARY]
  or    VARCHAR(length) [BINARY]
  or    DATE
  or    TIME
  or    TIMESTAMP
  or    DATETIME
  or    TINYBLOB
  or    BLOB
  or    MEDIUMBLOB
  or    LONGBLOB
  or    TINYTEXT
  or    TEXT
  or    MEDIUMTEXT
  or    LONGTEXT
  or    ENUM(value1,value2,value3,...)
  or    SET(value1,value2,value3,...)

index_col_name:
        col_name [(length)]

reference_definition:
        REFERENCES tbl_name [(index_col_name,...)]
                   [MATCH FULL | MATCH PARTIAL]
                   [ON DELETE reference_option]
                   [ON UPDATE reference_option]

reference_option:
        RESTRICT | CASCADE | SET NULL | NO ACTION | SET DEFAULT

table_options:
	TYPE = @{BDB | HEAP | ISAM | InnoDB | MERGE | MRG_MYISAM | MYISAM @}
or	AUTO_INCREMENT = #
or	AVG_ROW_LENGTH = #
or	CHECKSUM = @{0 | 1@}
or	COMMENT = "string"
or	MAX_ROWS = #
or	MIN_ROWS = #
or	PACK_KEYS = @{0 | 1 | DEFAULT@}
or	PASSWORD = "string"
or	DELAY_KEY_WRITE = @{0 | 1@}
or      ROW_FORMAT= @{ default | dynamic | fixed | compressed @}
or	RAID_TYPE= @{1 | STRIPED | RAID0 @} RAID_CHUNKS=#  RAID_CHUNKSIZE=#
or	UNION = (table_name,[table_name...])
or	INSERT_METHOD= @{NO | FIRST | LAST @}
or      DATA DIRECTORY="absolute path to directory"
or      INDEX DIRECTORY="absolute path to directory"

select_statement:
	[IGNORE | REPLACE] SELECT ...  (Some legal select statement)
@end example

@code{CREATE TABLE}
creates a table with the given name in the current database.  Rules for
allowable table names are given in @ref{Legal names}.  An error occurs if
there is no current database or if the table already exists.

In MySQL Version 3.22 or later, the table name can be specified as
@code{db_name.tbl_name}.  This works whether there is a current
database.

In MySQL Version 3.23, you can use the @code{TEMPORARY} keyword when
you create a table.  A temporary table will automatically be deleted if a
connection dies and the name is per connection.  This means that two different
connections can both use the same temporary table name without conflicting
with each other or with an existing table of the same name. (The existing table
is hidden until the temporary table is deleted.)

In MySQL Version 3.23 or later, you can use the keywords
@code{IF NOT EXISTS} so that an error does not occur if the table already
exists.  Note that there is no verification that the table structures are
identical.

Each table @code{tbl_name} is represented by some files in the database
directory. In the case of MyISAM-type tables you will get:

@multitable @columnfractions .20 .30
@item @strong{File} @tab @strong{Purpose}
@item @code{tbl_name.frm} @tab Table definition (form) file
@item @code{tbl_name.MYD} @tab Datafile
@item @code{tbl_name.MYI} @tab Index file
@end multitable

For more information on the properties of the various column types, see
@ref{Column types}:

@itemize @bullet
@item
If neither @code{NULL} nor @code{NOT NULL} is specified, the column
is treated as though @code{NULL} had been specified.

@item
An integer column may have the additional attribute @code{AUTO_INCREMENT}.
When you insert a value of @code{NULL} (recommended) or @code{0} into an
@code{AUTO_INCREMENT} column, the column is set to @code{value+1}, where
@code{value} is the largest value for the column currently in the table.
@code{AUTO_INCREMENT} sequences begin with @code{1}.
@xref{mysql_insert_id, , @code{mysql_insert_id()}}.

If you delete the row containing the maximum value for an
@code{AUTO_INCREMENT} column, the value will be reused with an
@code{ISAM}, or @code{BDB} table but not with a
@code{MyISAM} or @code{InnoDB} table.  If you delete all rows in the table
with @code{DELETE FROM table_name} (without a @code{WHERE}) in
@code{AUTOCOMMIT} mode, the sequence starts over for all table types.

@strong{Note}: there can be only one @code{AUTO_INCREMENT} column per
table, and it must be indexed. MySQL Version 3.23 will also only
work properly if the @code{AUTO_INCREMENT} column only has positive
values. Inserting a negative number is regarded as inserting a very large
positive number.  This is done to avoid precision problems when
numbers 'wrap' over from positive to negative and also to ensure that one
doesn't accidentally get an @code{AUTO_INCREMENT} column that contains 0.

In MyISAM and BDB tables you can specify @code{AUTO_INCREMENT} secondary
column in a multi-column key.  @xref{example-AUTO_INCREMENT}.

@cindex ODBC compatibility
@cindex compatibility, with ODBC
To make MySQL compatible with some ODBC applications, you can find
the last inserted row with the following query:

@example
SELECT * FROM tbl_name WHERE auto_col IS NULL
@end example

@item
@code{NULL} values are handled differently for @code{TIMESTAMP} columns than
for other column types.  You cannot store a literal @code{NULL} in a
@code{TIMESTAMP} column; setting the column to @code{NULL} sets it to the
current date and time.  Because @code{TIMESTAMP} columns behave this way, the
@code{NULL} and @code{NOT NULL} attributes do not apply in the normal way and
are ignored if you specify them.

On the other hand, to make it easier for MySQL clients to use
@code{TIMESTAMP} columns, the server reports that such columns may be
assigned @code{NULL} values (which is true), even though @code{TIMESTAMP}
never actually will contain a @code{NULL} value.  You can see this when you
use @code{DESCRIBE tbl_name} to get a description of your table.

Note that setting a @code{TIMESTAMP} column to @code{0} is not the same
as setting it to @code{NULL}, because @code{0} is a valid @code{TIMESTAMP}
value.

@item
@cindex default values
If no @code{DEFAULT} value is specified for a column, MySQL
automatically assigns one.

If the column may take @code{NULL} as a value, the default value is
@code{NULL}.

If the column is declared as @code{NOT NULL}, the default value depends on
the column type:

@itemize @minus
@item
For numeric types other than those declared with the @code{AUTO_INCREMENT}
attribute, the default is @code{0}.  For an @code{AUTO_INCREMENT} column, the
default value is the next value in the sequence.

@item
For date and time types other than @code{TIMESTAMP}, the default is the
appropriate zero value for the type.  For the first @code{TIMESTAMP}
column in a table, the default value is the current date and time.
@xref{Date and time types}.

@item
For string types other than @code{ENUM}, the default value is the empty
string.  For @code{ENUM}, the default is the first enumeration value (if
you haven't explicitly specified another default value with the
@code{DEFAULT} directive).
@end itemize

Default values must be constants. This means, for example, that you cannot
set the default for a date column to be the value of a function such as
@code{NOW()} or @code{CURRENT_DATE}.

@item
@code{KEY} is a synonym for @code{INDEX}.

@item
In MySQL, a @code{UNIQUE} key can have only distinct values. An
error occurs if you try to add a new row with a key that matches an existing
row.

@item
@tindex PRIMARY KEY
A @code{PRIMARY KEY} is a unique @code{KEY} with the extra constraint
that all key columns must be defined as @code{NOT NULL}.  In MySQL
the key is named @code{PRIMARY}. A table can have only one @code{PRIMARY KEY}.
If you don't have a @code{PRIMARY KEY} and some applications ask for the
@code{PRIMARY KEY} in your tables, MySQL will return the first
@code{UNIQUE} key, which doesn't have any @code{NULL} columns, as the
@code{PRIMARY KEY}.

@item
A @code{PRIMARY KEY} can be a multiple-column index.  However, you cannot
create a multiple-column index using the @code{PRIMARY KEY} key attibute in a
column specification.  Doing so will mark only that single column as primary.
You must use the @code{PRIMARY KEY(index_col_name, ...)} syntax.

@item
If the @code{PRIMARY} or @code{UNIQUE} key consists of only one column and this
is of type integer, you can also refer to it as @code{_rowid}
(new in Version 3.23.11).

@item
If you don't assign a name to an index, the index will be assigned the same
name as the first @code{index_col_name}, with an optional suffix (@code{_2},
@code{_3}, @code{...}) to make it unique.  You can see index names for a
table using @code{SHOW INDEX FROM tbl_name}.
@xref{SHOW, , @code{SHOW}}.

@item
@cindex @code{NULL} values, and indexes
@cindex indexes, and @code{NULL} values
Only the @code{MyISAM}, @code{InnoDB}, and @code{BDB} table types support indexes on columns that can have
@code{NULL} values. In other cases you must declare such columns
@code{NOT NULL} or an error results.

@item
With @code{col_name(length)} syntax, you can specify an index that
uses only a part of a @code{CHAR} or @code{VARCHAR} column. This can
make the index file much smaller.
@xref{Indexes}.

@item
@cindex @code{BLOB} columns, indexing
@cindex indexes, and @code{BLOB} columns
@cindex @code{TEXT} columns, indexing
@cindex indexes, and @code{TEXT} columns
Only the @code{MyISAM} table type supports indexing on @code{BLOB} and
@code{TEXT} columns.  When putting an index on a @code{BLOB} or @code{TEXT}
column you MUST always specify the length of the index:
@example
CREATE TABLE test (blob_col BLOB, INDEX(blob_col(10)));
@end example

@item
When you use @code{ORDER BY} or @code{GROUP BY} with a @code{TEXT} or
@code{BLOB} column, only the first @code{max_sort_length} bytes are used.
@xref{BLOB,  , @code{BLOB}}.

@item
In MySQL Version 3.23.23 or later, you can also create special
@strong{FULLTEXT} indexes. They are used for full-text search. Only the
@code{MyISAM} table type supports @code{FULLTEXT} indexes. They can be created
only from @code{VARCHAR} and @code{TEXT} columns.
Indexing always happens over the entire column, partial indexing is not
supported. See @ref{Fulltext Search} for details of operation.

@item
The @code{FOREIGN KEY}, @code{CHECK}, and @code{REFERENCES} clauses don't
actually do anything.  The syntax for them is provided only for compatibility,
to make it easier to port code from other SQL servers and to run applications
that create tables with references.
@xref{Differences from ANSI}.

@item
Each @code{NULL} column takes one bit extra, rounded up to the nearest byte.

@item
The maximum record length in bytes can be calculated as follows:

@example
row length = 1
             + (sum of column lengths)
             + (number of NULL columns + 7)/8
             + (number of variable-length columns)
@end example

@item
The @code{table_options} and @code{SELECT} options are only
implemented in MySQL Version 3.23 and above.

The different table types are:

@multitable @columnfractions .25 .70
@item @strong{Table type} @tab @strong{Description}
@item BDB or Berkeley_db @tab Transaction-safe tables with page locking. @xref{BDB}.
@item HEAP @tab The data for this table is only stored in memory. @xref{HEAP}.
@item ISAM @tab The original table handler. @xref{ISAM}.
@item InnoDB @tab Transaction-safe tables with row locking. @xref{InnoDB}.
@item MERGE @tab A collection of MyISAM tables used as one table. @xref{MERGE}.
@item MRG_MyISAM @tab An alias for MERGE tables
@item MyISAM @tab The new binary portable table handler that is replacing ISAM. @xref{MyISAM}.
@end multitable
@xref{Table types}.

If a table type is specified, and that particular type is not available,
MySQL will choose the closest table type to the one that you have
specified.  For example, if @code{TYPE=BDB} is specified, and that distribution
of MySQL does not support @code{BDB} tables, the table will be created
as @code{MyISAM} instead.

The other table options are used to optimise the behavior of the
table. In most cases, you don't have to specify any of them.
The options work for all table types, if not otherwise indicated:

@multitable @columnfractions .25 .70
@item @strong{Option} @tab @strong{Description}
@item @code{AUTO_INCREMENT} @tab The next @code{AUTO_INCREMENT} value you want to set for your table (MyISAM).
@item @code{AVG_ROW_LENGTH} @tab An approximation of the average row length for your table. You only need to set this for large tables with variable size records.
@item @code{CHECKSUM} @tab Set this to 1 if you want MySQL to maintain a checksum for all rows (makes the table a little slower to update but makes it easier to find corrupted tables) (MyISAM).
@item @code{COMMENT} @tab A 60-character comment for your table.
@item @code{MAX_ROWS} @tab Max number of rows you plan to store in the table.
@item @code{MIN_ROWS} @tab Minimum number of rows you plan to store in the table.
@item @code{PACK_KEYS} @tab Set this to 1 if you want to have a smaller index. This usually makes updates slower and reads faster (MyISAM, ISAM). Setting this to 0 will disable all packing of keys. Setting this to @code{DEFAULT} (MySQL 4.0) will tell the table handler to only pack long @code{CHAR}/@code{VARCHAR} columns.
@item @code{PASSWORD} @tab Encrypt the @file{.frm} file with a password.  This option doesn't do anything in the standard MySQL version.
@item @code{DELAY_KEY_WRITE} @tab Set this to 1 if want to delay key table updates until the table is closed (MyISAM).
@item @code{ROW_FORMAT} @tab Defines how the rows should be stored. Currently this option only works with MyISAM tables, which supports the @code{DYNAMIC} and @code{FIXED} row formats. @xref{MyISAM table formats}.
@end multitable

When you use a @code{MyISAM} table, MySQL uses the product of
@code{max_rows * avg_row_length} to decide how big the resulting table
will be.  If you don't specify any of the above options, the maximum size
for a table will be 4G (or 2G if your operating systems only supports 2G
tables). The reason for this is just to keep down the pointer sizes
to make the index smaller and faster if you don't really need big files.

If you don't use @code{PACK_KEYS}, the default is to only pack strings,
not numbers.  If you use @code{PACK_KEYS=1}, numbers will be packed as well.

When packing binary number keys, MySQL will use prefix compression.
This means that you will only get a big benefit of this if you have
many numbers that are the same.  Prefix compression means that every
key needs one extra byte to indicate how many bytes of the previous key are
the same for the next key (note that the pointer to the row is stored
in high-byte-first-order directly after the key, to improve
compression).  This means that if you have many equal keys on two rows
in a row, all following 'same' keys will usually only take 2 bytes
(including the pointer to the row).  Compare this to the ordinary case
where the following keys will take storage_size_for_key +
pointer_size (usually 4).  On the other hand, if all keys are
totally different, you will lose 1 byte per key, if the key isn't a
key that can have @code{NULL} values. (In this case the packed key length will
be stored in the same byte that is used to mark if a key is @code{NULL}.)

@item
If you specify a @code{SELECT} after the @code{CREATE} statement,
MySQL will create new fields for all elements in the
@code{SELECT}.  For example:

@example
mysql> CREATE TABLE test (a INT NOT NULL AUTO_INCREMENT,
    ->        PRIMARY KEY (a), KEY(b))
    ->        TYPE=MyISAM SELECT b,c FROM test2;
@end example

This will create a @code{MyISAM} table with three columns, a, b, and c.
Notice that the columns from the @code{SELECT} statement are appended to
the right side of the table, not overlapped onto it.  Take the following
example:

@example
mysql> SELECT * FROM foo;
+---+
| n |
+---+
| 1 |
+---+

mysql> CREATE TABLE bar (m INT) SELECT n FROM foo;
Query OK, 1 row affected (0.02 sec)
Records: 1  Duplicates: 0  Warnings: 0

mysql> SELECT * FROM bar;
+------+---+
| m    | n |
+------+---+
| NULL | 1 |
+------+---+
1 row in set (0.00 sec)
@end example

For each row in table @code{foo}, a row is inserted in @code{bar} with
the values from @code{foo} and default values for the new columns.

@code{CREATE TABLE ... SELECT} will not automatically create any indexes
for you.  This is done intentionally to make the command as flexible as
possible.  If you want to have indexes in the created table, you should
specify these before the @code{SELECT} statement:

@example
mysql> CREATE TABLE bar (UNIQUE (n)) SELECT n FROM foo;
@end example

If any errors occur while copying the data to the table, it will
automatically be deleted.

To ensure that the update log/binary log can be used to re-create the
original tables, MySQL will not allow concurrent inserts during
@code{CREATE TABLE ... SELECT}.
@item
The @code{RAID_TYPE} option will help you to break the 2G/4G limit for
the MyISAM datafile (not the index file) on operating systems that
don't support big files. Note that this option is not recommended for
filesystem that supports big files!

You can get more speed from the I/O bottleneck by putting @code{RAID}
directories on different physical disks. @code{RAID_TYPE} will work on
any OS, as long as you have configured MySQL with @code{--with-raid}.
For now the only allowed @code{RAID_TYPE} is @code{STRIPED} (@code{1}
and @code{RAID0} are aliases for this).

If you specify @code{RAID_TYPE=STRIPED} for a @code{MyISAM} table,
@code{MyISAM} will create @code{RAID_CHUNKS} subdirectories named 00,
01, 02 in the database directory.  In each of these directories
@code{MyISAM} will create a @code{table_name.MYD}.  When writing data
to the datafile, the @code{RAID} handler will map the first
@code{RAID_CHUNKSIZE} *1024 bytes to the first file, the next
@code{RAID_CHUNKSIZE} *1024 bytes to the next file and so on.
@item
@code{UNION} is used when you want to use a collection of identical
tables as one. This only works with @code{MERGE} tables.
@xref{MERGE, , @code{MERGE}}.

For the moment you need to have @strong{select}, @strong{update}, and
@strong{delete} privileges on the tables you map to a @code{MERGE} table.
All mapped tables must be in the same database as the @code{MERGE} table.

@item
If you want to insert data in a @code{MERGE} table, you have to specify with
@code{INSERT_METHOD} into with table the row should be inserted.
@xref{MERGE}.  This option was introduced in MySQL 4.0.0.

@item
In the created table the @code{PRIMARY} key will be placed first, followed
by all @code{UNIQUE} keys and then the normal keys.  This helps the
MySQL optimiser to prioritise which key to use and also more quickly
detect duplicated @code{UNIQUE} keys.

@item
By using @code{DATA DIRECTORY="directory"} or @code{INDEX
DIRECTORY="directory"} you can specify where the table handler should
put it's table and index files.  Note that the directory should be a full
path to the directory (not relative path).

This only works for @code{MyISAM} tables in @code{MySQL} 4.0, when you
are not using the @code{--skip-symlink} option. @xref{Symbolic links to
tables}.
@end itemize


@node Silent column changes,  , CREATE TABLE, CREATE TABLE
@subsubsection Silent Column Specification Changes

@cindex silent column changes

In some cases, MySQL silently changes a column specification from
that given in a @code{CREATE TABLE} statement.  (This may also occur with
@code{ALTER TABLE}.):

@itemize @bullet
@item
@code{VARCHAR} columns with a length less than four are changed to
@code{CHAR}.

@item
If any column in a table has a variable length, the entire row is
variable-length as a result.  Therefore, if a table contains any
variable-length columns (@code{VARCHAR}, @code{TEXT}, or @code{BLOB}),
all @code{CHAR} columns longer than three characters are changed to
@code{VARCHAR} columns.  This doesn't affect how you use the columns in
any way; in MySQL, @code{VARCHAR} is just a different way to
store characters.  MySQL performs this conversion because it
saves space and makes table operations faster.  @xref{Table types}.

@item
@code{TIMESTAMP} display sizes must be even and in the range from 2 to 14.
If you specify a display size of 0 or greater than 14, the size is coerced
to 14.  Odd-valued sizes in the range from 1 to 13 are coerced
to the next higher even number.

@item
You cannot store a literal @code{NULL} in a @code{TIMESTAMP} column; setting
it to @code{NULL} sets it to the current date and time.  Because
@code{TIMESTAMP} columns behave this way, the @code{NULL} and @code{NOT NULL}
attributes do not apply in the normal way and are ignored if you specify
them.  @code{DESCRIBE tbl_name} always reports that a @code{TIMESTAMP}
column may be assigned @code{NULL} values.

@item
MySQL maps certain column types used by other SQL database vendors
to MySQL types.  @xref{Other-vendor column types}.
@end itemize

If you want to see whether MySQL used a column type other
than the one you specified, issue a @code{DESCRIBE tbl_name} statement after
creating or altering your table.

@cindex @code{myisampack}
Certain other column type changes may occur if you compress a table
using @code{myisampack}. @xref{Compressed format}.


@node ALTER TABLE, RENAME TABLE, CREATE TABLE, Data Definition
@subsection @code{ALTER TABLE} Syntax

@findex ALTER TABLE

@example
ALTER [IGNORE] TABLE tbl_name alter_spec [, alter_spec ...]

alter_specification:
        ADD [COLUMN] create_definition [FIRST | AFTER column_name ]
  or    ADD [COLUMN] (create_definition, create_definition,...)
  or    ADD INDEX [index_name] (index_col_name,...)
  or    ADD PRIMARY KEY (index_col_name,...)
  or    ADD UNIQUE [index_name] (index_col_name,...)
  or    ADD FULLTEXT [index_name] (index_col_name,...)
  or	ADD [CONSTRAINT symbol] FOREIGN KEY index_name (index_col_name,...)
            [reference_definition]
  or    ALTER [COLUMN] col_name @{SET DEFAULT literal | DROP DEFAULT@}
  or    CHANGE [COLUMN] old_col_name create_definition
               [FIRST | AFTER column_name]
  or    MODIFY [COLUMN] create_definition [FIRST | AFTER column_name]
  or    DROP [COLUMN] col_name
  or    DROP PRIMARY KEY
  or    DROP INDEX index_name
  or    DISABLE KEYS
  or    ENABLE KEYS
  or    RENAME [TO] new_tbl_name
  or    ORDER BY col
  or    table_options
@end example

@code{ALTER TABLE} allows you to change the structure of an existing table.
For example, you can add or delete columns, create or destroy indexes, change
the type of existing columns, or rename columns or the table itself.  You can
also change the comment for the table and type of the table.
@xref{CREATE TABLE, , @code{CREATE TABLE}}.

If you use @code{ALTER TABLE} to change a column specification but
@code{DESCRIBE tbl_name} indicates that your column was not changed, it is
possible that MySQL ignored your modification for one of the reasons
described in @ref{Silent column changes}.  For example, if you try to change
a @code{VARCHAR} column to @code{CHAR}, MySQL will still use
@code{VARCHAR} if the table contains other variable-length columns.

@code{ALTER TABLE} works by making a temporary copy of the original table.
The alteration is performed on the copy, then the original table is
deleted and the new one is renamed. This is done in such a way that
all updates are automatically redirected to the new table without
any failed updates. While @code{ALTER TABLE} is executing, the original
table is readable by other clients. Updates and writes to the table
are stalled until the new table is ready.

Note that if you use any other option to @code{ALTER TABLE} than
@code{RENAME}, MySQL will always create a temporary table, even
if the data wouldn't strictly need to be copied (like when you change the
name of a column). We plan to fix this in the future, but as one doesn't
normally do @code{ALTER TABLE} that often this isn't that high on our TODO.
For MyISAM tables, you can speed up the index recreation part (which is the
slowest part of the recreation process) by setting the
@code{myisam_sort_buffer_size} variable to a high value.

@itemize @bullet
@item
To use @code{ALTER TABLE}, you need @strong{alter}, @strong{insert},
and @strong{create} privileges on the table.

@item
@code{IGNORE} is a MySQL extension to ANSI SQL92.
It controls how @code{ALTER TABLE} works if there are duplicates on
unique keys in the new table.
If @code{IGNORE} isn't specified, the copy is aborted and rolled back.
If @code{IGNORE} is specified, then for rows with duplicates on a unique
key, only the first row is used; the others are deleted.

@item
You can issue multiple @code{ADD}, @code{ALTER}, @code{DROP}, and
@code{CHANGE} clauses in a single @code{ALTER TABLE} statement. This is a
MySQL extension to ANSI SQL92, which allows only one of each clause
per @code{ALTER TABLE} statement.

@item
@code{CHANGE col_name}, @code{DROP col_name}, and @code{DROP
INDEX} are MySQL extensions to ANSI SQL92.

@item
@code{MODIFY} is an Oracle extension to @code{ALTER TABLE}.

@item
The optional word @code{COLUMN} is a pure noise word and can be omitted.

@item
If you use @code{ALTER TABLE tbl_name RENAME TO new_name} without any other
options, MySQL simply renames the files that correspond to the table
@code{tbl_name}.  There is no need to create the temporary table.
@xref{RENAME TABLE,, @code{RENAME TABLE}}.

@item
@code{create_definition} clauses use the same syntax for @code{ADD} and
@code{CHANGE} as for @code{CREATE TABLE}.  Note that this syntax includes
the column name, not just the column type.
@xref{CREATE TABLE, , @code{CREATE TABLE}}.

@item
You can rename a column using a @code{CHANGE old_col_name create_definition}
clause.  To do so, specify the old and new column names and the type that
the column currently has.  For example, to rename an @code{INTEGER} column
from @code{a} to @code{b}, you can do this:

@example
mysql> ALTER TABLE t1 CHANGE a b INTEGER;
@end example

If you want to change a column's type but not the name, @code{CHANGE}
syntax still requires two column names even if they are the same.  For
example:

@example
mysql> ALTER TABLE t1 CHANGE b b BIGINT NOT NULL;
@end example

However, as of MySQL Version 3.22.16a, you can also use @code{MODIFY}
to change a column's type without renaming it:

@example
mysql> ALTER TABLE t1 MODIFY b BIGINT NOT NULL;
@end example

@item
If you use @code{CHANGE} or @code{MODIFY} to shorten a column for which
an index exists on part of the column (for instance, if you have an index
on the first 10 characters of a @code{VARCHAR} column), you cannot make
the column shorter than the number of characters that are indexed.

@item
When you change a column type using @code{CHANGE} or @code{MODIFY},
MySQL tries to convert data to the new type as well as possible.

@item
In MySQL Version 3.22 or later, you can use @code{FIRST} or
@code{ADD ... AFTER col_name} to add a column at a specific position within
a table row. The default is to add the column last.

@findex ALTER COLUMN
@item
@code{ALTER COLUMN} specifies a new default value for a column
or removes the old default value.
If the old default is removed and the column can be @code{NULL}, the new
default is @code{NULL}. If the column cannot be @code{NULL}, MySQL
assigns a default value, as described in
@ref{CREATE TABLE, , @code{CREATE TABLE}}.

@findex DROP INDEX
@item
@code{DROP INDEX} removes an index. This is a MySQL extension to
ANSI SQL92. @xref{DROP INDEX}.

@item
If columns are dropped from a table, the columns are also removed from any
index of which they are a part.  If all columns that make up an index are
dropped, the index is dropped as well.

@item
If a table contains only one column, the column cannot be dropped.
If what you intend is to remove the table, use @code{DROP TABLE} instead.

@findex DROP PRIMARY KEY
@item
@code{DROP PRIMARY KEY} drops the primary index. If no such
index exists, it drops the first @code{UNIQUE} index in the table.
(MySQL marks the first @code{UNIQUE} key as the @code{PRIMARY KEY}
if no @code{PRIMARY KEY} was specified explicitly.)

@findex UNIQUE
@findex PRIMARY KEY
If you add a @code{UNIQUE INDEX} or @code{PRIMARY KEY} to a table, this
is stored before any not @code{UNIQUE} index so that MySQL can detect
duplicate keys as early as possible.

@findex ORDER BY
@item
@code{ORDER BY} allows you to create the new table with the rows in a
specific order.  Note that the table will not remain in this order after
inserts and deletes.  In some cases, it may make sorting easier for
MySQL if the table is in order by the column that you wish to
order it by later.  This option is mainly useful when you know that you
are mostly going to query the rows in a certain order; by using this
option after big changes to the table, you may be able to get higher
performance.

@findex ALTER TABLE
@item
If you use @code{ALTER TABLE} on a @code{MyISAM} table, all non-unique
indexes are created in a separate batch (like in @code{REPAIR}).
This should make @code{ALTER TABLE} much faster when you have many indexes.

@item
Since @strong{MySQL 4.0} the above feature can be activated explicitly.
@code{ALTER TABLE ... DISABLE KEYS} makes MySQL to stop updating
non-unique indexes for @code{MyISAM} table.
@code{ALTER TABLE ... ENABLE KEYS} then should be used to recreate missing
indexes. As MySQL does it with special algorithm which is much
faster then inserting keys one by one, disabling keys could give a
considerable speedup on bulk inserts.

@item
@findex mysql_info()
With the C API function @code{mysql_info()}, you can find out how many
records were copied, and (when @code{IGNORE} is used) how many records were
deleted due to duplication of unique key values.

@item
@cindex foreign keys
@cindex references
The @code{FOREIGN KEY}, @code{CHECK}, and @code{REFERENCES} clauses don't
actually do anything.  The syntax for them is provided only for compatibility,
to make it easier to port code from other SQL servers and to run applications
that create tables with references.
@xref{Differences from ANSI}.
@end itemize

Here is an example that shows some of the uses of @code{ALTER TABLE}.  We
begin with a table @code{t1} that is created as shown here:

@example
mysql> CREATE TABLE t1 (a INTEGER,b CHAR(10));
@end example

To rename the table from @code{t1} to @code{t2}:

@example
mysql> ALTER TABLE t1 RENAME t2;
@end example

To change column @code{a} from @code{INTEGER} to @code{TINYINT NOT NULL}
(leaving the name the same), and to change column @code{b} from
@code{CHAR(10)} to @code{CHAR(20)} as well as renaming it from @code{b} to
@code{c}:

@example
mysql> ALTER TABLE t2 MODIFY a TINYINT NOT NULL, CHANGE b c CHAR(20);
@end example

To add a new @code{TIMESTAMP} column named @code{d}:

@example
mysql> ALTER TABLE t2 ADD d TIMESTAMP;
@end example

To add an index on column @code{d}, and make column @code{a} the primary key:

@example
mysql> ALTER TABLE t2 ADD INDEX (d), ADD PRIMARY KEY (a);
@end example

To remove column @code{c}:

@example
mysql> ALTER TABLE t2 DROP COLUMN c;
@end example

To add a new @code{AUTO_INCREMENT} integer column named @code{c}:

@example
mysql> ALTER TABLE t2 ADD c INT UNSIGNED NOT NULL AUTO_INCREMENT,
           ADD INDEX (c);
@end example

Note that we indexed @code{c}, because @code{AUTO_INCREMENT} columns must be
indexed, and also that we declare @code{c} as @code{NOT NULL}, because
indexed columns cannot be @code{NULL}.

When you add an @code{AUTO_INCREMENT} column, column values are filled in
with sequence numbers for you automatically.  You can set the first
sequence number by executing @code{SET INSERT_ID=#} before
@code{ALTER TABLE} or using the @code{AUTO_INCREMENT = #} table option.
@xref{SET OPTION}.

With MyISAM tables, if you don't change the @code{AUTO_INCREMENT}
column, the sequence number will not be affected. If you drop an
@code{AUTO_INCREMENT} column and then add another @code{AUTO_INCREMENT}
column, the numbers will start from 1 again.

@xref{ALTER TABLE problems}.


@node RENAME TABLE, DROP TABLE, ALTER TABLE, Data Definition
@subsection @code{RENAME TABLE} Syntax

@findex RENAME TABLE

@example
RENAME TABLE tbl_name TO new_tbl_name[, tbl_name2 TO new_tbl_name2,...]
@end example

The rename is done atomically, which means that no other thread can
access any of the tables while the rename is running. This makes it
possible to replace a table with an empty one:

@example
CREATE TABLE new_table (...);
RENAME TABLE old_table TO backup_table, new_table TO old_table;
@end example

The rename is done from left to right, which means that if you want to
swap two tables names, you have to:

@example
RENAME TABLE old_table    TO backup_table,
             new_table    TO old_table,
             backup_table TO new_table;
@end example

As long as two databases are on the same disk you can also rename
from one database to another:

@example
RENAME TABLE current_db.tbl_name TO other_db.tbl_name;
@end example

When you execute @code{RENAME}, you can't have any locked tables or
active transactions.  You must also have the @strong{alter} and @strong{drop}
privileges on the original table, and the @strong{create} and @strong{insert}
privileges on the new table.

If MySQL encounters any errors in a multiple-table rename, it
will do a reverse rename for all renamed tables to get everything back
to the original state.

@code{RENAME TABLE} was added in MySQL 3.23.23.

@node DROP TABLE, CREATE INDEX, RENAME TABLE, Data Definition
@subsection @code{DROP TABLE} Syntax

@findex DROP TABLE

@example
DROP TABLE [IF EXISTS] tbl_name [, tbl_name,...] [RESTRICT | CASCADE]
@end example

@code{DROP TABLE} removes one or more tables. All table data and the table
definition are @emph{removed}, so @strong{be careful} with this command!

In MySQL Version 3.22 or later, you can use the keywords
@code{IF EXISTS} to prevent an error from occurring for tables that don't
exist.

@code{RESTRICT} and @code{CASCADE} are allowed to make porting easier.
For the moment they don't do anything.

@strong{Note}: @code{DROP TABLE} will
automatically commit current active transaction.


@node CREATE INDEX, DROP INDEX, DROP TABLE, Data Definition
@subsection @code{CREATE INDEX} Syntax

@findex CREATE INDEX

@cindex indexes
@cindex indexes, multi-part
@cindex multi-part index

@example
CREATE [UNIQUE|FULLTEXT] INDEX index_name
       ON tbl_name (col_name[(length)],... )
@end example

The @code{CREATE INDEX} statement doesn't do anything in MySQL prior
to Version 3.22.  In Version 3.22 or later, @code{CREATE INDEX} is mapped to an
@code{ALTER TABLE} statement to create indexes.
@xref{ALTER TABLE, , @code{ALTER TABLE}}.

Normally, you create all indexes on a table at the time the table itself
is created with @code{CREATE TABLE}.
@xref{CREATE TABLE, , @code{CREATE TABLE}}.
@code{CREATE INDEX} allows you to add indexes to existing tables.

A column list of the form @code{(col1,col2,...)} creates a multiple-column
index.  Index values are formed by concatenating the values of the given
columns.

For @code{CHAR} and @code{VARCHAR} columns, indexes can be created that
use only part of a column, using @code{col_name(length)} syntax.  (On
@code{BLOB} and @code{TEXT} columns the length is required.) The
statement shown here creates an index using the first 10 characters of
the @code{name} column:

@example
mysql> CREATE INDEX part_of_name ON customer (name(10));
@end example

Because most names usually differ in the first 10 characters, this index should
not be much slower than an index created from the entire @code{name} column.
Also, using partial columns for indexes can make the index file much smaller,
which could save a lot of disk space and might also speed up @code{INSERT}
operations!

Note that you can only add an index on a column that can have @code{NULL}
values or on a @code{BLOB}/@code{TEXT} column if you are using
MySQL Version 3.23.2 or newer and are using the @code{MyISAM}
table type.

For more information about how MySQL uses indexes, see
@ref{MySQL indexes, , MySQL indexes}.

@code{FULLTEXT} indexes can index only @code{VARCHAR} and
@code{TEXT} columns, and only in @code{MyISAM} tables. @code{FULLTEXT} indexes
are available in MySQL Version 3.23.23 and later.
@ref{Fulltext Search}.


@node DROP INDEX,  , CREATE INDEX, Data Definition
@subsection @code{DROP INDEX} Syntax

@findex DROP INDEX

@example
DROP INDEX index_name ON tbl_name
@end example

@code{DROP INDEX} drops the index named @code{index_name} from the table
@code{tbl_name}.  @code{DROP INDEX} doesn't do anything in MySQL
prior to Version 3.22.  In Version 3.22 or later, @code{DROP INDEX} is mapped to an
@code{ALTER TABLE} statement to drop the index.
@xref{ALTER TABLE, , @code{ALTER TABLE}}.


@node Basic User Commands, Transactional Commands, Data Definition, Reference
@section Basic MySQL User Utility Commands

@menu
* USE::                         @code{USE} Syntax
* DESCRIBE::                    @code{DESCRIBE} Syntax (Get Information About Columns)
@end menu


@node USE, DESCRIBE, Basic User Commands, Basic User Commands
@subsection @code{USE} Syntax

@findex USE

@example
USE db_name
@end example

The @code{USE db_name} statement tells MySQL to use the @code{db_name}
database as the default database for subsequent queries.  The database remains
current until the end of the session or until another @code{USE} statement
is issued:

@example
mysql> USE db1;
mysql> SELECT COUNT(*) FROM mytable;      # selects from db1.mytable
mysql> USE db2;
mysql> SELECT COUNT(*) FROM mytable;      # selects from db2.mytable
@end example

Making a particular database current by means of the @code{USE} statement
does not preclude you from accessing tables in other databases.  The following example
accesses the @code{author} table from the @code{db1} database and the
@code{editor} table from the @code{db2} database:

@example
mysql> USE db1;
mysql> SELECT author_name,editor_name FROM author,db2.editor
    ->        WHERE author.editor_id = db2.editor.editor_id;
@end example

@cindex Sybase compatibility
@cindex compatibility, with Sybase
The @code{USE} statement is provided for Sybase compatibility.


@node DESCRIBE,  , USE, Basic User Commands
@subsection @code{DESCRIBE} Syntax (Get Information About Columns)

@findex DESC
@findex DESCRIBE

@example
@{DESCRIBE | DESC@} tbl_name @{col_name | wild@}
@end example

@code{DESCRIBE} is a shortcut for @code{SHOW COLUMNS FROM}.
@xref{SHOW DATABASE INFO}.

@code{DESCRIBE} provides information about a table's columns.  @code{col_name}
may be a column name or a string containing the SQL @samp{%} and @samp{_}
wildcard characters.

If the column types are different from what you expect them to be based on a
@code{CREATE TABLE} statement, note that MySQL sometimes
changes column types.  @xref{Silent column changes}.

@cindex Oracle compatibility
@cindex compatibility, with Oracle
This statement is provided for Oracle compatibility.

The @code{SHOW} statement provides similar information.
@xref{SHOW, , @code{SHOW}}.


@node Transactional Commands, Fulltext Search, Basic User Commands, Reference
@section MySQL Transactional and Locking Commands

@menu
* COMMIT::                      @code{BEGIN/COMMIT/ROLLBACK} Syntax
* LOCK TABLES::                 @code{LOCK TABLES/UNLOCK TABLES} Syntax
* SET TRANSACTION::             @code{SET TRANSACTION} Syntax
@end menu


@node COMMIT, LOCK TABLES, Transactional Commands, Transactional Commands
@subsection @code{BEGIN/COMMIT/ROLLBACK} Syntax

@findex BEGIN
@findex COMMIT
@findex ROLLBACK

By default, MySQL runs in @code{autocommit} mode. This means that
as soon as you execute an update, MySQL will store the update on
disk.

If you are using transactions safe tables (like @code{InnoDB},
@code{BDB}, you can put MySQL into
non-@code{autocommit} mode with the following command:

@example
SET AUTOCOMMIT=0
@end example

After this you must use @code{COMMIT} to store your changes to disk or
@code{ROLLBACK} if you want to ignore the changes you have made since
the beginning of your transaction.

If you want to switch from @code{AUTOCOMMIT} mode for one series of
statements, you can use the @code{BEGIN} or @code{BEGIN WORK} statement:

@example
BEGIN;
SELECT @@A:=SUM(salary) FROM table1 WHERE type=1;
UPDATE table2 SET summmary=@@A WHERE type=1;
COMMIT;
@end example

Note that if you are using non-transaction-safe tables, the changes will be
stored at once, independent of the status of the @code{autocommit} mode.

If you do a @code{ROLLBACK} when you have updated a non-transactional
table you will get an error (@code{ER_WARNING_NOT_COMPLETE_ROLLBACK}) as
a warning.  All transactional safe tables will be restored but any
non-transactional table will not change.

If you are using @code{BEGIN} or @code{SET AUTOCOMMIT=0}, you
should use the MySQL binary log for backups instead of the
older update log.  Transactions are stored in the binary log
in one chunk, upon @code{COMMIT}, to ensure that transactions which are
rolled back are not stored.  @xref{Binary log}.

The following commands automatically end a transaction (as if you had done
a @code{COMMIT} before executing the command):

@multitable @columnfractions .25 .25 .25
@item @strong{Command}      @tab @strong{Command}   @tab @strong{Command}
@item @code{ALTER TABLE}    @tab @code{BEGIN}       @tab @code{CREATE INDEX}
@item @code{DROP DATABASE}  @tab @code{DROP TABLE}  @tab @code{RENAME TABLE}
@item @code{TRUNCATE}       @tab                    @tab
@end multitable

You can change the isolation level for transactions with
@code{SET TRANSACTION ISOLATION LEVEL ...}. @xref{SET TRANSACTION}.


@node LOCK TABLES, SET TRANSACTION, COMMIT, Transactional Commands
@subsection @code{LOCK TABLES/UNLOCK TABLES} Syntax

@findex LOCK TABLES
@findex UNLOCK TABLES

@example
LOCK TABLES tbl_name [AS alias] @{READ | [READ LOCAL] | [LOW_PRIORITY] WRITE@}
            [, tbl_name @{READ | [LOW_PRIORITY] WRITE@} ...]
...
UNLOCK TABLES
@end example

@code{LOCK TABLES} locks tables for the current thread.  @code{UNLOCK
TABLES} releases any locks held by the current thread.  All tables that
are locked by the current thread are automatically unlocked when the
thread issues another @code{LOCK TABLES}, or when the connection to the
server is closed.

The main reasons to use @code{LOCK TABLES} are for emulating transactions
or getting more speed when updating tables.  This is explained in more
detail later.

If a thread obtains a @code{READ} lock on a table, that thread (and all other
threads) can only read from the table. If a thread obtains a @code{WRITE}
lock on a table, then only the thread holding the lock can @code{READ} from
or @code{WRITE} to the table.  Other threads are blocked.

The difference between @code{READ LOCAL} and @code{READ} is that
@code{READ LOCAL} allows non-conflicting @code{INSERT} statements to
execute while the lock is held.  This can't however be used if you are
going to manipulate the database files outside MySQL while you
hold the lock.

When you use @code{LOCK TABLES}, you must lock all tables that you are
going to use and you must use the same alias that you are going to use
in your queries!  If you are using a table multiple times in a query
(with aliases), you must get a lock for each alias!

@code{WRITE} locks normally have higher priority than @code{READ} locks, to
ensure that updates are processed as soon as possible. This means that if one
thread obtains a @code{READ} lock and then another thread requests a
@code{WRITE} lock, subsequent @code{READ} lock requests will wait until the
@code{WRITE} thread has gotten the lock and released it.  You can use
@code{LOW_PRIORITY WRITE} locks to allow other threads to obtain @code{READ}
locks while the thread is waiting for the @code{WRITE} lock. You should only
use @code{LOW_PRIORITY WRITE} locks if you are sure that there will
eventually be a time when no threads will have a @code{READ} lock.

@code{LOCK TABLES} works as follows:
@enumerate
@item
Sort all tables to be locked in a internally defined order (from the
user standpoint the order is undefined).
@item
If a table is locked with a read and a write lock, put the write lock
before the read lock.
@item
Lock one table at a time until the thread gets all locks.
@end enumerate

This policy ensures that table locking is deadlock free. There is
however other things one needs to be aware of with this schema:

If you are using a @code{LOW_PRIORITY_WRITE} lock for a table, this
means only that MySQL will wait for this particlar lock until
there is no threads that wants a @code{READ} lock.  When the thread has
got the @code{WRITE} lock and is waiting to get the lock for the next
table in the lock table list, all other threads will wait for the
@code{WRITE} lock to be released.  If this becomes a serious problem
with your application, you should consider converting some of your
tables to transactions safe tables.

You can safely kill a thread that is waiting for a table lock with
@code{KILL}.  @xref{KILL}.

Note that you should @strong{not} lock any tables that you are using with
@code{INSERT DELAYED}.  This is because that in this case the @code{INSERT}
is done by a separate thread.

Normally, you don't have to lock tables, as all single @code{UPDATE} statements
are atomic; no other thread can interfere with any other currently executing
SQL statement. There are a few cases when you would like to lock tables
anyway:

@itemize @bullet
@item
If you are going to run many operations on a bunch of tables, it's much
faster to lock the tables you are going to use.  The downside is, of course,
that no other thread can update a @code{READ}-locked table and no other
thread can read a @code{WRITE}-locked table.

The reason some things are faster under @code{LOCK TABLES} is that
MySQL will not flush the key cache for the locked tables until
@code{UNLOCK TABLES} is called (normally the key cache is flushed after
each SQL statement). This speeds up inserting/updateing/deletes on
@code{MyISAM} tables.
@item
If you are using a table handler in MySQL that doesn't support
transactions, you must use @code{LOCK TABLES} if you want to ensure that
no other thread comes between a @code{SELECT} and an @code{UPDATE}. The
example shown here requires @code{LOCK TABLES} in order to execute safely:

@example
mysql> LOCK TABLES trans READ, customer WRITE;
mysql> SELECT SUM(value) FROM trans WHERE customer_id=some_id;
mysql> UPDATE customer SET total_value=sum_from_previous_statement
    ->        WHERE customer_id=some_id;
mysql> UNLOCK TABLES;
@end example

Without @code{LOCK TABLES}, there is a chance that another thread might
insert a new row in the @code{trans} table between execution of the
@code{SELECT} and @code{UPDATE} statements.
@end itemize

By using incremental updates (@code{UPDATE customer SET
value=value+new_value}) or the @code{LAST_INSERT_ID()} function, you can
avoid using @code{LOCK TABLES} in many cases.

You can also solve some cases by using the user-level lock functions
@code{GET_LOCK()} and @code{RELEASE_LOCK()}.  These locks are saved in a hash
table in the server and implemented with @code{pthread_mutex_lock()} and
@code{pthread_mutex_unlock()} for high speed.
@xref{Miscellaneous functions}.

See @ref{Internal locking}, for more information on locking policy.

You can lock all tables in all databases with read locks with the
@code{FLUSH TABLES WITH READ LOCK} command. @xref{FLUSH}. This is very
convenient way to get backups if you have a filesystem, like Veritas,
that can take snapshots in time.

@strong{NOTE}: @code{LOCK TABLES} is not transaction-safe and will
automatically commit any active transactions before attempting to lock the
tables.


@node SET TRANSACTION,  , LOCK TABLES, Transactional Commands
@subsection @code{SET TRANSACTION} Syntax

@findex SET TRANSACTION
@findex ISOLATION LEVEL

@example
SET [GLOBAL | SESSION] TRANSACTION ISOLATION LEVEL
@{ READ UNCOMMITTED | READ COMMITTED | REPEATABLE READ | SERIALIZABLE @}
@end example

Sets the transaction isolation level for the global, whole session or
the next transaction.

The default behavior is to set the isolation level for the next (not
started) transaction.  If you use the @code{GLOBAL} keyword, the statement
sets the default transaction level globally for all new connections
created from that point on.  You will need the @strong{process}
privilege to do do this.  Using the @code{SESSION} keyword sets the
default transaction level for all future transactions performed on the
current connection.

You can set the default global isolation level for @code{mysqld} with
@code{--transaction-isolation=...}. @xref{Command-line options}.

@node Fulltext Search, Query Cache, Transactional Commands, Reference
@section MySQL Full-text Search

@cindex searching, full-text
@cindex full-text search
@cindex FULLTEXT

As of Version 3.23.23, MySQL has support for full-text indexing
and searching.  Full-text indexes in MySQL are an index of type
@code{FULLTEXT}.  @code{FULLTEXT} indexes can be created from @code{VARCHAR}
and @code{TEXT} columns at @code{CREATE TABLE} time or added later with
@code{ALTER TABLE} or @code{CREATE INDEX}.  For large datasets, it will be
much faster to load your data into a table that has no @code{FULLTEXT}
index, then create the index with @code{ALTER TABLE} (or @code{CREATE
INDEX}).  Loading data into a table that already has a @code{FULLTEXT}
index will be slower.

Full-text searching is performed with the @code{MATCH()} function.

@example
mysql> CREATE TABLE articles (
    ->   id INT UNSIGNED AUTO_INCREMENT NOT NULL PRIMARY KEY,
    ->   title VARCHAR(200),
    ->   body TEXT,
    ->   FULLTEXT (title,body)
    -> );
Query OK, 0 rows affected (0.00 sec)

mysql> INSERT INTO articles VALUES
    -> (0,'MySQL Tutorial', 'DBMS stands for DataBase ...'),
    -> (0,'How To Use MySQL Efficiently', 'After you went through a ...'),
    -> (0,'Optimising MySQL','In this tutorial we will show ...'),
    -> (0,'1001 MySQL Trick','1. Never run mysqld as root. 2. ...'),
    -> (0,'MySQL vs. YourSQL', 'In the following database comparison ...'),
    -> (0,'MySQL Security', 'When configured properly, MySQL ...');
Query OK, 6 rows affected (0.00 sec)
Records: 6  Duplicates: 0  Warnings: 0

mysql> SELECT * FROM articles
    ->          WHERE MATCH (title,body) AGAINST ('database');
+----+-------------------+------------------------------------------+
| id | title             | body                                     |
+----+-------------------+------------------------------------------+
|  5 | MySQL vs. YourSQL | In the following database comparison ... |
|  1 | MySQL Tutorial    | DBMS stands for DataBase ...             |
+----+-------------------+------------------------------------------+
2 rows in set (0.00 sec)
@end example

The @code{MATCH()} function performs a natural language search for a string
against a text collection (a set of of one or more columns included in
a @code{FULLTEXT} index).  The search string is given as the argument to
@code{AGAINST()}.  The search is performed in case-insensitive fashion.
For every row in the table, @code{MATCH()} returns a relevance value,
that is, a similarity measure between the search string and the text in
that row in the columns named in the @code{MATCH()} list.

When @code{MATCH()} is used in a @code{WHERE} clause (see example above)
the rows returned are automatically sorted with highest relevance first.
Relevance values are non-negative floating-point numbers.  Zero relevance
means no similarity.  Relevance is computed based on the number of words
in the row, the number of unique words in that row, the total number of
words in the collection, and the number of documents (rows) that contain
a particular word.

It is also possible to perform a boolean mode search.  This is explained
later in the section.

The preceding example is a basic illustration showing how to use the
@code{MATCH()} function. Rows are returned in order of decreasing
relevance.

The next example shows how to retrieve the relevance values explicitly.
As neither @code{WHERE} nor @code{ORDER BY} clauses are present, returned
rows are not ordered.

@example
mysql> SELECT id,MATCH (title,body) AGAINST ('Tutorial') FROM articles;
+----+-----------------------------------------+
| id | MATCH (title,body) AGAINST ('Tutorial') |
+----+-----------------------------------------+
|  1 |                        0.64840710366884 |
|  2 |                                       0 |
|  3 |                        0.66266459031789 |
|  4 |                                       0 |
|  5 |                                       0 |
|  6 |                                       0 |
+----+-----------------------------------------+
6 rows in set (0.00 sec)
@end example

The following example is more complex.  The query returns the relevance
and still sorts the rows in order of decreasing relevance. To achieve
this result, you should specify @code{MATCH()} twice. This will cause no
additional overhead, because the MySQL optimiser will notice that the
two @code{MATCH()} calls are identical and invoke the full-text search
code only once.

@example
mysql> SELECT id, body, MATCH (title,body) AGAINST
    -> ('Security implications of running MySQL as root') AS score
    -> FROM articles WHERE MATCH (title,body) AGAINST
    -> ('Security implications of running MySQL as root');
+----+-------------------------------------+-----------------+
| id | body                                | score           |
+----+-------------------------------------+-----------------+
|  4 | 1. Never run mysqld as root. 2. ... | 1.5055546709332 |
|  6 | When configured properly, MySQL ... |   1.31140957288 |
+----+-------------------------------------+-----------------+
2 rows in set (0.00 sec)
@end example

MySQL uses a very simple parser to split text into words.  A ``word''
is any sequence of characters consisting of letters, numbers, @samp{'},
and @samp{_}.  Any ``word'' that is present in the stopword list or is just
too short (3 characters or less) is ignored.

Every correct word in the collection and in the query is weighted
according to its significance in the query or collection.  This way, a
word that is present in many documents will have lower weight (and may
even have a zero weight), because it has lower semantic value in this
particular collection.  Otherwise, if the word is rare, it will receive a
higher weight.  The weights of the words are then combined to compute the
relevance of the row.

Such a technique works best with large collections (in fact, it was
carefully tuned this way).  For very small tables, word distribution
does not reflect adequately their semantic value, and this model
may sometimes produce bizarre results.

@example
mysql> SELECT * FROM articles WHERE MATCH (title,body) AGAINST ('MySQL');
Empty set (0.00 sec)
@end example

The search for the word @code{MySQL} produces no results in the above
example, because that word is present in more than half the rows.  As such,
it is effectively treated as a stopword (that is, a word with zero semantic
value).  This is the most desirable behavior -- a natural language query
should not return every second row from a 1GB table.

A word that matches half of rows in a table is less likely to locate relevant
documents.  In fact, it will most likely find plenty of irrelevant documents.
We all know this happens far too often when we are trying to find something on
the Internet with a search engine.  It is with this reasoning that such rows
have been assigned a low semantic value in @strong{this particular dataset}.

As of Version 4.0.1, MySQL can also perform boolean full-text searches using
the @code{IN BOOLEAN MODE} modifier.

@example
mysql> SELECT * FROM articles WHERE MATCH (title,body)
    ->     AGAINST ('+MySQL -YourSQL' IN BOOLEAN MODE);
+----+------------------------------+-------------------------------------+
| id | title                        | body                                |
+----+------------------------------+-------------------------------------+
|  1 | MySQL Tutorial               | DBMS stands for DataBase ...        |
|  2 | How To Use MySQL Efficiently | After you went through a ...        |
|  3 | Optimising MySQL             | In this tutorial we will show ...   |
|  4 | 1001 MySQL Trick             | 1. Never run mysqld as root. 2. ... |
|  6 | MySQL Security               | When configured properly, MySQL ... |
+----+------------------------------+-------------------------------------+
@end example

This query retrieved all the rows that contain the word @code{MySQL}
(note: the 50% threshold is not used), but that do @strong{not} contain
the word @code{YourSQL}.  Note that a boolean mode search does not
auto-magically sort rows in order of decreasing relevance.  You can
see this from result of the preceding query, where the row with the
highest relevance (the one that contains @code{MySQL} twice) is listed
last, not first.  A boolean full-text search can also work even without
a @code{FULLTEXT} index, although it would be @strong{slow}.

The boolean full-text search capability supports the following operators:

@table @code
@item +
A leading plus sign indicates that this word @strong{must be}
present in every row returned.
@item -
A leading minus sign indicates that this word @strong{must not be}
present in any row returned.
@item
By default (when neither plus nor minus is specified) the word is optional,
but the rows that contain it will be rated higher. This mimicks the
behaviour of @code{MATCH() ... AGAINST()} without the @code{IN BOOLEAN
MODE} modifier.
@item < >
These two operators are used to change a word's contribution to the
relevance value that is assigned to a row.  The @code{<} operator
decreases the contribution and the @code{>} operator increases it.
See the example below.
@item ( )
Parentheses are used to group words into subexpressions.
@item ~
A leading tilde acts as a negation operator, causing the word's
contribution to the row relevance to be negative. It's useful for marking
noise words. A row that contains such a word will be rated lower than
others, but will not be excluded altogether, as it would be with the
@code{-} operator.
@item *
An asterisk is the truncation operator. Unlike the other operators, it
should be @strong{appended} to the word, not prepended.
@item "
The phrase, that is enclosed in double quotes @code{"}, matches only
rows that contain this phrase @strong{literally, as it was typed}.
@end table

And here are some examples:

@table @code
@item apple banana
find rows that contain at least one of these words.
@item +apple +juice
... both words.
@item +apple macintosh
... word ``apple'', but rank it higher if it also contain ``macintosh''.
@item +apple -macintosh
... word ``apple'' but not ``macintosh''.
@item +apple +(>pie <strudel)
... ``apple'' and ``pie'', or ``apple'' and ``strudel'' (in any
order), but rank ``apple pie'' higher than ``apple strudel''.
@item apple*
... ``apple'', ``apples'', ``applesauce'', and ``applet''.
@item "some words"
... ``some words of wisdom'', but not ``some noise words''.
@end table

@menu
* Fulltext Restrictions::       Full-text Restrictions
* Fulltext Fine-tuning::        Fine-tuning MySQL Full-text Search
* Fulltext TODO::               Full-text Search TODO
@end menu

@node Fulltext Restrictions, Fulltext Fine-tuning, Fulltext Search, Fulltext Search
@subsection Full-text Restrictions

@itemize @bullet
@item
All parameters to the @code{MATCH()} function must be columns from the
same table that is part of the same @code{FULLTEXT} index, unless the
@code{MATCH()} is @code{IN BOOLEAN MODE}.
@item
The @code{MATCH()} column list must exactly match the column list in some
@code{FULLTEXT} index definition for the table, unless this @code{MATCH()}
is @code{IN BOOLEAN MODE}.
@item
The argument to @code{AGAINST()} must be a constant string.
@end itemize


@node Fulltext Fine-tuning, Fulltext TODO, Fulltext Restrictions, Fulltext Search
@subsection Fine-tuning MySQL Full-text Search

Unfortunately, full-text search has few user-tunable parameters yet,
although adding some is very high on the TODO. If you have a
MySQL source distribution (@pxref{Installing source}), you can
exert more control over full-text searching behavior.

Note that full-text search was carefully tuned for the best searching
effectiveness.  Modifying the default behavior will, in most cases,
only make the search results worse.  Do not alter the MySQL sources
unless you know what you are doing!

@itemize @bullet

@item
The minimum length of words to be indexed is defined by the MySQL
variable @code{ft_min_word_length}. @xref{SHOW VARIABLES}.
Change it to the value you prefer, and rebuild your @code{FULLTEXT} indexes.
(This variable is only available from MySQL version 4.0.)

@item
The stopword list is defined in @file{myisam/ft_static.c}
Modify it to your taste, recompile MySQL, and rebuild
your @code{FULLTEXT} indexes.

@item
The 50% threshold is determined by the particular weighting scheme chosen.
To disable it, change the following line in @file{myisam/ftdefs.h}:
@example
#define GWS_IN_USE GWS_PROB
@end example
To:
@example
#define GWS_IN_USE GWS_FREQ
@end example
Then recompile MySQL.
There is no need to rebuild the indexes in this case.
@strong{Note}: by doing this you @strong{severely} decrease MySQL's ability
to provide adequate relevance values for the @code{MATCH()} function.
If you really need to search for such common words, it would be better to
search using @code{IN BOOLEAN MODE} instead, which does not observe the 50%
threshold.

@item
Sometimes the search engine maintainer would like to change the operators used
for boolean full-text searches. These are defined by the
@code{ft_boolean_syntax} variable. @xref{SHOW VARIABLES}.
Still, this variable is read-only, its value is set in
@file{myisam/ft_static.c}.

@end itemize

@node Fulltext TODO,  , Fulltext Fine-tuning, Fulltext Search
@subsection Full-text Search TODO

@itemize @bullet
@item Make all operations with @code{FULLTEXT} index @strong{faster}.
@item Proximity operators
@item Support for "always-index words". They could be any strings
the user wants to treat as words, examples are "C++", "AS/400", "TCP/IP", etc.
@item Support for full-text search in @code{MERGE} tables.
@item Support for multi-byte charsets.
@item Make stopword list to depend of the language of the data.
@item Stemming (dependent of the language of the data, of course).
@item Generic user-suppliable UDF preparser.
@item Make the model more flexible (by adding some adjustable
parameters to @code{FULLTEXT} in @code{CREATE/ALTER TABLE}).
@end itemize


@node Query Cache,  , Fulltext Search, Reference
@section MySQL Query Cache

@cindex Query Cache
@cindex @code{SELECT}, Query Cache

From version 4.0.1, @code{MySQL server} features a @code{Query Cache}.
When in use, the query cache stores the text of a @code{SELECT} query
together with the corresponding result that is sent to a client.
If another identical query is received, the server can then retrieve
the results from the query cache rather than parsing and executing the
same query again.

The query cache is extremely useful in an environment where (some)
tables don't change very often and you have a lot of identical queries.
This is a typical situation for many web servers that use a lot of
dynamic content.

Following are some performance data for the query cache
(We got these by running the MySQL benchmark suite on a Linux Alpha
2x500 MHz with 2GB RAM and a 64MB query cache):

@itemize @bullet
@item
If you want to disable the query cache code set @code{query_cache_size=0}.
By disabling the query cache code there is no noticeable overhead.
(query cache can be excluded from code with help of configure option
@code{--without-query-cache}) 
@item
If all of the queries you're preforming are simple (such as selecting a
row from a table with one row); but still differ so that the queries can
not be cached, the overhead for having the query cache active is 13%.
This could be regarded as the worst case scenario. However, in real life,
queries are much more complicated than our simple example so the overhead
is normally significantly lower.
@item
Searches after one row in a one row table is 238% faster.
This can be regarded as close to the minimum speedup to be expected for
a query that is cached.
@end itemize

@menu
* Query Cache How::             How The Query Cache Operates
* Query Cache Configuration::   Query Cache Configuration
* Query Cache in SELECT::       Query Cache Options in @code{SELECT}
* Query Cache Status and Maintenance::  Query Cache Status and Maintenance
@end menu


@node Query Cache How, Query Cache Configuration, Query Cache, Query Cache
@subsection How The Query Cache Operates

Queries are compared before parsing, thus

@example
SELECT * FROM TABLE
@end example

and

@example
Select * from table
@end example

are regarded as different queries for query cache, so queries need
to be exactly the same (byte for byte) to be seen as identical.
In addition, a query may be seen as different if for instance one
client is using a new communication protocol format or another
character set than another client.

Queries that uses different databases, uses different protocol versions
or the uses different default character sets are considered different
queries and cached separately.

The cache does work for @code{SELECT CALC_ROWS ...} and
@code{SELECT FOUND_ROWS() ...} type queries because the number of
found rows is also stored in the cache.

If a table changes (@code{INSERT}, @code{UPDATE}, @code{DELETE},
@code{TRUNCATE}, @code{ALTER} or @code{DROP TABLE|DATABASE}),
then all cached queries that used this table (possibly through a
@code{MRG_MyISAM} table!) become invalid and are removed from the cache.

Transactional @code{InnoDB} tables that have been changed will be invalidated
when a @code{COMMIT} is performed.

A query cannot be cached if it contains one of the functions:
@multitable @columnfractions .33 .33 .34
@item @strong{Function}
 @tab @strong{Function}
 @tab @strong{Function}
@item @code{User-Defined Functions}
 @tab @code{CONNECTION_ID}
 @tab @code{FOUND_ROWS}
@item @code{GET_LOCK}
 @tab @code{RELEASE_LOCK}
 @tab @code{LOAD_FILE}
@item @code{MASTER_POS_WAIT}
 @tab @code{NOW}
 @tab @code{SYSDATE}
@item @code{CURRENT_TIMESTAMP}
 @tab @code{CURDATE}
 @tab @code{CURRENT_DATE}
@item @code{CURTIME}
 @tab @code{CURRENT_TIME}
 @tab @code{DATABASE}
@item @code{ENCRYPT} (with one parameter)
 @tab @code{LAST_INSERT_ID}
 @tab @code{RAND}
@item @code{UNIX_TIMESTAMP} (without parameters)
 @tab @code{USER}
 @tab @code{BENCHMARK}
@end multitable

Nor can a query be cached if it contains user variables,
if it is of the form @code{SELECT ... IN SHARE MODE} or
of the form @code{SELECT * FROM AUTOINCREMENT_FIELD IS NULL}
(to retrieve last insert id - ODBC work around).

However, @code{FOUND ROWS()} will return the correct value,
even if the preceding query was fetched from the cache.

Queries that don't use any tables or if the user has a column privilege for
any of the involved tables are not cached.

Before a query is fetched from the query cache, MySQL will check that
the user has SELECT privilege to all the involved databases and
tables. If this is not the case, the cached result will not be used.

@node Query Cache Configuration, Query Cache in SELECT, Query Cache How, Query Cache
@subsection Query Cache Configuration

The query cache adds a few @code{MySQL} system variables for
@code{mysqld} which may be set in a configuration file, on the
command-line when starting @code{mysqld}.

@itemize
@item @code{query_cache_limit}
Don't cache results that are bigger than this. (Default 1M).

@item @code{query_cache_size}
The memory allocated to store results from old queries.
If this is 0, the query cache is disabled (default).

@item @code{query_cache_startup_type}
This may be set (only numeric) to
@multitable @columnfractions .10 .75
@item @strong{Option} @tab @strong{Description}
@item 0 @tab (OFF, don't cache or retrieve results)
@item 1 @tab (ON, cache all results except @code{SELECT SQL_NO_CACHE ...} queries)
@item 2 @tab (DEMAND, cache only @code{SELECT SQL_CACHE ...} queries)
@end multitable
@end itemize


Inside a thread (connection), the behaviour of the query cache can be
changed from the default. The syntax is as follows:

@code{SQL_QUERY_CACHE_TYPE = OFF | ON | DEMAND}
@code{SQL_QUERY_CACHE_TYPE = 0   | 1  | 2}

@multitable @columnfractions .20 .70
@item @strong{Option} @tab @strong{Description}
@item 0 or OFF @tab Don't cache or retrieve results.
@item 1 or ON @tab Cache all results except @code{SELECT SQL_NO_CACHE ...} queries.
@item 2 or DEMAND @tab Cache only @code{SELECT SQL_CACHE ...} queries.
@end multitable

By default @code{SQL_QUERY_CACHE_TYPE} depends on the value of
@code{query_cache_startup_type} when the thread was created.


@node Query Cache in SELECT, Query Cache Status and Maintenance, Query Cache Configuration, Query Cache
@subsection Query Cache Options in @code{SELECT}

There are two possible query cache related parameters that may be
specified in a @code{SELECT} query:

@findex SQL_CACHE
@findex SQL_NO_CACHE

@multitable @columnfractions .20 .80
@item @strong{Option} @tab @strong{Description}
@item @code{SQL_CACHE}
 @tab If @code{SQL_QUERY_CACHE_TYPE} is @code{DEMAND}, allow the query to be cached.
 If @code{SQL_QUERY_CACHE_TYPE} is @code{ON}, this is the default.
 If @code{SQL_QUERY_CACHE_TYPE} is @code{OFF}, do nothing.
@item @code{SQL_NO_CACHE}
 @tab Make this query non-cachable, don't allow this query to be stored in the cache.
@end multitable


@node Query Cache Status and Maintenance,  , Query Cache in SELECT, Query Cache
@subsection Query Cache Status and Maintenance

With the @code{FLUSH QUERY CACHE} command you can defragment the query
cache to better utilise its memory. This command will not remove any
queries from the cache.
@code{FLUSH TABLES} also flushes the query cache.

The @code{RESET QUERY CACHE} command removes all query results from the
query cache.

You can monitor query cache performance in @code{SHOW STATUS}:

@multitable @columnfractions .30 .50
@item @strong{Variable} @tab @strong{Description}
@item @code{Qcache_queries_in_cache}
 @tab Number of queries registered in the cache.
@item @code{Qcache_inserts}
 @tab Number of queries added to the cache.
@item @code{Qcache_hits}
 @tab Number of cache hits.
@item @code{Qcache_not_cached}
 @tab Number of non-cached queries
 (not cachable, or due to @code{SQL_QUERY_CACHE_TYPE}).
@item @code{Qcache_free_memory}
 @tab Amount of free memory for query cache.
@item @code{Qcache_total_blocks}
 @tab Total number of blocks in query cache.
@item @code{Qcache_free_blocks}
 @tab Number of free memory blocks in query cache.
@end multitable

Total number of queries =
@code{Qcache_inserts} + @code{Qcache_hits} + @code{Qcache_not_cached}.

The query cache uses variable length blocks, so @code{Qcache_total_blocks}
and @code{Qcache_free_blocks} may indicate query cache memory fragmentation.
After @code{FLUSH QUERY CACHE} only a single (big) free block remains.

Note: Every query needs a minimum of two blocks (one for the query text
and one or more for the query results). Also, every table that is used
by a query needs one block, but if two or more queries use same table
only one block needs to be allocated.


@node Table types, Clients, Reference, Top
@chapter MySQL Table Types

@cindex table types, choosing
@cindex @code{BDB} table type
@cindex @code{Berkeley_db} table type
@cindex @code{HEAP} table type
@cindex @code{ISAM} table type
@cindex @code{InnoDB} table type
@cindex @code{MERGE} table type
@cindex MySQL table types
@cindex @code{MyISAM} table type
@cindex types, of tables

As of MySQL Version 3.23.6, you can choose between three basic
table formats (@code{ISAM}, @code{HEAP} and @code{MyISAM}.  Newer
MySQL may support additional table type (@code{InnoDB},
or @code{BDB}), depending on how you compile it.

When you create a new table, you can tell MySQL which table
type it should use for the table.  MySQL will always create a
@file{.frm} file to hold the table and column definitions.  Depending on
the table type, the index and data will be stored in other files.

Note that to use @code{InnoDB} tables you have to use at least
the @code{innodb_data_file_path} startup option. @xref{InnoDB start}.

The default table type in MySQL is @code{MyISAM}. If you are
trying to use a table type that is not compiled-in or activated,
MySQL will instead create a table of type @code{MyISAM}.  This
is a very useful feature when you want to copy tables between different
SQL servers that supports different table types (like copying tables to
a slave that is optimised for speed by not having transactional tables).
This automatic table changing can however also be very confusing for new
MySQL users. We plan to fix this by introducing warnings in
MySQL 4.0 and giving a warning when a table type is automatically
changed.

You can convert tables between different types with the @code{ALTER
TABLE} statement. @xref{ALTER TABLE, , @code{ALTER TABLE}}.

Note that MySQL supports two different kinds of
tables: transaction-safe tables (@code{InnoDB} and @code{BDB})
and not transaction-safe tables (@code{HEAP}, @code{ISAM},
@code{MERGE}, and @code{MyISAM}).

Advantages of transaction-safe tables (TST):

@itemize @bullet
@item
Safer. Even if MySQL crashes or you get hardware problems, you
can get your data back, either by automatic recovery or from a backup
+ the transaction log.
@item
You can combine many statements and accept these all in one go with
the @code{COMMIT} command.
@item
You can execute @code{ROLLBACK} to ignore your changes (if you are not
running in auto commit mode).
@item
If an update fails, all your changes will be restored. (With NTST tables all
changes that have taken place are permanent)
@end itemize

Advantages of not transaction-safe tables (NTST):

@itemize @bullet
@item
Much faster as there is no transaction overhead.
@item
Will use less disk space as there is no overhead of transactions.
@item
Will use less memory to do updates.
@end itemize

You can combine TST and NTST tables in the same statements to get the best
of both worlds.

@menu
* MyISAM::                      MyISAM tables
* MERGE::                       MERGE tables
* ISAM::                        ISAM tables
* HEAP::                        HEAP tables
* InnoDB::                      InnoDB tables
* BDB::                         BDB or Berkeley_db tables
@end menu


@node MyISAM, MERGE, Table types, Table types
@section MyISAM Tables

@code{MyISAM} is the default table type in MySQL Version 3.23. It's
based on the @code{ISAM} code and has a lot of useful extensions.

The index is stored in a file with the @file{.MYI} (MYIndex) extension,
and the data is stored in a file with the @file{.MYD} (MYData) extension.
You can check/repair @code{MyISAM} tables with the @code{myisamchk}
utility. @xref{Crash recovery}. You can compress @code{MyISAM} tables with
@code{myisampack} to take up much less space.
@xref{myisampack, , @code{myisampack}}.

The following is new in @code{MyISAM}:

@itemize @bullet
@item
There is a flag in the @code{MyISAM} file that indicates whether
the table was closed correctly.  If @code{mysqld} is started with
@code{--myisam-recover}, @code{MyISAM} tables will automatically be
checked and/or repaired on open if the table wasn't closed properly.
@item
You can @code{INSERT} new rows in a table that doesn't have free blocks
in the middle of the datafile, at the same time other threads are
reading from the table (concurrent insert).  An free block can come from
an update of a dynamic length row with much data to a row with less data
or when deleting rows. When all free blocks are used up, all future
inserts will be concurrent again.
@item
Support for big files (63-bit) on filesystems/operating systems that
support big files.
@item
All data is stored with the low byte first. This makes the data machine
and OS independent. The only requirement is that the machine uses
two's-complement signed integers (as every machine for the last 20 years
has) and IEEE floating-point format (also totally dominant among
mainstream machines). The only area of machines that may not support
binary compatibility are embedded systems (because they sometimes have
peculiar processors).

There is no big speed penalty in storing data low byte first; the bytes
in a table row is normally unaligned and it doesn't take that much more
power to read an unaligned byte in order than in reverse order.  The
actual fetch-column-value code is also not time critical compared to
other code.
@item
All number keys are stored with high byte first to give better index
compression.
@item
Internal handling of one @code{AUTO_INCREMENT} column. @code{MyISAM}
will automatically update this on @code{INSERT/UPDATE}. The
@code{AUTO_INCREMENT} value can be reset with @code{myisamchk}. This
will make @code{AUTO_INCREMENT} columns faster (at least 10 %) and old
numbers will not be reused as with the old @code{ISAM}. Note that when an
@code{AUTO_INCREMENT} is defined on the end of a multi-part-key the old
behavior is still present.
@item
When inserted in sorted order (as when you are using an @code{AUTO_INCREMENT}
column) the key tree will be split so that the high node only contains one
key. This will improve the space utilisation in the key tree.
@item
@code{BLOB} and @code{TEXT} columns can be indexed.
@item
@code{NULL} values are allowed in indexed columns.  This takes 0-1
bytes/key.
@item
Maximum key length is 500 bytes by default (can be changed by
recompiling).  In cases of keys longer than 250 bytes, a bigger key
block size than the default of 1024 bytes is used for this key.
@item
Maximum number of keys/table is 32 as default. This can be enlarged to 64
without having to recompile @code{myisamchk}.
@item
@code{myisamchk} will mark tables as checked if one runs it with
@code{--update-state}. @code{myisamchk --fast} will only check those
tables that don't have this mark.
@item
@code{myisamchk -a} stores statistics for key parts (and not only for
whole keys as in @code{ISAM}).
@item
Dynamic size rows will now be much less fragmented when mixing deletes
with updates and inserts.  This is done by automatically combining adjacent
deleted blocks and by extending blocks if the next block is deleted.
@item
@code{myisampack} can pack @code{BLOB} and @code{VARCHAR} columns.
@item
You can use put the datafile and index file on different directories
to get more speed (with the @code{DATA/INDEX DIRECTORY="path"} option to
@code{CREATE TABLE}). @xref{CREATE TABLE}.
@end itemize

@code{MyISAM} also supports the following things, which MySQL
will be able to use in the near future:

@itemize @bullet
@item
Support for a true @code{VARCHAR} type; a @code{VARCHAR} column starts
with a length stored in 2 bytes.
@item
Tables with @code{VARCHAR} may have fixed or dynamic record length.
@item
@code{VARCHAR} and @code{CHAR} may be up to 64K.
All key segments have their own language definition. This will enable
MySQL to have different language definitions per column.
@item
A hashed computed index can be used for @code{UNIQUE}. This will allow
you to have @code{UNIQUE} on any combination of columns in a table. (You
can't search on a @code{UNIQUE} computed index, however.)
@end itemize

Note that index files are usually much smaller with @code{MyISAM} than with
@code{ISAM}. This means that @code{MyISAM} will normally use less
system resources than @code{ISAM}, but will need more CPU time when inserting
data into a compressed index.

The following options to @code{mysqld} can be used to change the behavior of
@code{MyISAM} tables. @xref{SHOW VARIABLES}.

@multitable @columnfractions .40 .60
@item @strong{Option} @tab @strong{Description}
@item @code{--myisam-recover=#} @tab Automatic recovery of crashed tables.
@item @code{-O myisam_sort_buffer_size=#} @tab Buffer used when recovering tables.
@item @code{--delay-key-write-for-all-tables} @tab Don't flush key buffers between writes for any MyISAM table
@item @code{-O myisam_max_extra_sort_file_size=#} @tab Used to help MySQL to decide when to use the slow but safe key cache index create method. @strong{Note} that this parameter is given in megabytes!
@item @code{-O myisam_max_sort_file_size=#} @tab Don't use the fast sort index method to created index if the temporary file would get bigger than this.  @strong{Note} that this paramter is given in megabytes!
@item @code{-O myisam_bulk_insert_tree_size=#} @tab Size of tree cache used in bulk insert optimisation. @strong{Note} that this is a limit @strong{per thread}!
@end multitable

The automatic recovery is activated if you start @code{mysqld} with
@code{--myisam-recover=#}. @xref{Command-line options}.
On open, the table is checked if it's marked as crashed or if the open
count variable for the table is not 0 and you are running with
@code{--skip-locking}.  If either of the above is true the following happens.

@itemize @bullet
@item
The table is checked for errors.
@item
If we found an error, try to do a fast repair (with sorting and without
re-creating the datafile) of the table.
@item
If the repair fails because of an error in the datafile (for example a
duplicate key error), we try again, but this time we re-create the datafile.
@item
If the repair fails, retry once more with the old repair option method
(write row by row without sorting) which should be able to repair any
type of error with little disk requirements..
@end itemize

If the recover wouldn't be able to recover all rows from a previous
completed statement and you didn't specify @code{FORCE} as an option to
@code{myisam-recover}, then the automatic repair will abort with an error
message in the error file:

@example
Error: Couldn't repair table: test.g00pages
@end example

If you in this case had used the @code{FORCE} option you would instead have got
a warning in the error file:

@example
Warning: Found 344 of 354 rows when repairing ./test/g00pages
@end example

Note that if you run automatic recover with the @code{BACKUP} option,
you should have a cron script that automatically moves file with names
like @file{tablename-datetime.BAK} from the database directories to a
backup media.

@xref{Command-line options}.

@menu
* Key space::                   Space needed for keys
* MyISAM table formats::        MyISAM table formats
* MyISAM table problems::       MyISAM table problems.
@end menu


@node Key space, MyISAM table formats, MyISAM, MyISAM
@subsection Space Needed for Keys

@cindex key space, MyISAM

MySQL can support different index types, but the normal type is
ISAM or MyISAM.  These use a B-tree index, and you can roughly calculate
the size for the index file as @code{(key_length+4)/0.67}, summed over
all keys.  (This is for the worst case when all keys are inserted in
sorted order and we don't have any compressed keys.)

String indexes are space compressed. If the first index part is a
string, it will also be prefix compressed.  Space compression makes the
index file smaller than the above figures if the string column has a lot
of trailing space or is a @code{VARCHAR} column that is not always used
to the full length. Prefix compression is used on keys that start
with a string.  Prefix compression helps if there are many strings
with an identical prefix.

In @code{MyISAM} tables, you can also prefix compress numbers by specifying
@code{PACK_KEYS=1} when you create the table.  This helps when you have
many integer keys that have an identical prefix when the numbers are stored
high-byte first.


@node MyISAM table formats, MyISAM table problems, Key space, MyISAM
@subsection MyISAM Table Formats

@code{MyISAM} supports 3 different table types. Two of them are chosen
automatically depending on the type of columns you are using. The third,
compressed tables, can only be created with the @code{myisampack} tool.

When you @code{CREATE} or @code{ALTER} a table you can for tables that
doesn't have @code{BLOB}'s force the table format to @code{DYNAMIC} or
@code{FIXED} with the @code{ROW_FORMAT=#} table option.  In the future
you will be able to compress/decompress tables by specifying
@code{ROW_FORMAT=compressed | default} to @code{ALTER TABLE}.
@xref{CREATE TABLE}.

@menu
* Static format::               Static (Fixed-length) table characteristics
* Dynamic format::              Dynamic table characteristics
* Compressed format::           Compressed table characteristics
@end menu


@node Static format, Dynamic format, MyISAM table formats, MyISAM table formats
@subsubsection Static (Fixed-length) Table Characteristics

This is the default format. It's used when the table contains no
@code{VARCHAR}, @code{BLOB}, or @code{TEXT} columns.

This format is the simplest and most secure format. It is also the
fastest of the on-disk formats. The speed comes from the easy way data
can be found on disk. When looking up something with an index and static
format it is very simple. Just multiply the row number by the row length.

Also, when scanning a table it is very easy to read a constant number of
records with each disk read.

The security is evidenced if your computer crashes when writing to a
fixed-size MyISAM file, in which case @code{myisamchk} can easily figure out where each
row starts and ends. So it can usually reclaim all records except the
partially written one. Note that in MySQL all indexes can always be
reconstructed:

@itemize @bullet
@item
All @code{CHAR}, @code{NUMERIC}, and @code{DECIMAL} columns are space-padded
to the column width.
@item
Very quick.
@item
Easy to cache.
@item
Easy to reconstruct after a crash, because records are located in fixed
positions.
@item
Doesn't have to be reorganised (with @code{myisamchk}) unless a huge number of
records are deleted and you want to return free disk space to the operating
system.
@item
Usually requires more disk space than dynamic tables.
@end itemize


@node Dynamic format, Compressed format, Static format, MyISAM table formats
@subsubsection Dynamic Table Characteristics

@cindex dynamic table characteristics
@cindex tables, dynamic

This format is used if the table contains any @code{VARCHAR}, @code{BLOB},
or @code{TEXT} columns or if the table was created with
@code{ROW_FORMAT=dynamic}.

This format is a little more complex because each row has to have a header
that says how long it is. One record can also end up at more than one
location when it is made longer at an update.

@cindex tables, defragment
You can use @code{OPTIMIZE table} or @code{myisamchk} to defragment a
table. If you have static data that you access/change a lot in the same
table as some @code{VARCHAR} or @code{BLOB} columns, it might be a good
idea to move the dynamic columns to other tables just to avoid
fragmentation:

@itemize @bullet
@item
All string columns are dynamic (except those with a length less than 4).
@item
Each record is preceded by a bitmap indicating which columns are empty
(@code{''}) for string columns, or zero for numeric columns. (This isn't
the same as columns containing @code{NULL} values.) If a string column
has a length of zero after removal of trailing spaces, or a numeric
column has a value of zero, it is marked in the bit map and not saved to
disk.  Non-empty strings are saved as a length byte plus the string
contents.
@item
Usually takes much less disk space than fixed-length tables.
@item
Each record uses only as much space as is required. If a record becomes
larger, it is split into as many pieces as are required.  This results in record
fragmentation.
@item
If you update a row with information that extends the row length, the
row will be fragmented.  In this case, you may have to run @code{myisamchk
-r} from time to time to get better performance.  Use @code{myisamchk -ei
tbl_name} for some statistics.
@item
Not as easy to reconstruct after a crash, because a record may be fragmented
into many pieces and a link (fragment) may be missing.
@item
The expected row length for dynamic sized records is:

@example
3
+ (number of columns + 7) / 8
+ (number of char columns)
+ packed size of numeric columns
+ length of strings
+ (number of NULL columns + 7) / 8
@end example

There is a penalty of 6 bytes for each link. A dynamic record is linked
whenever an update causes an enlargement of the record. Each new link will be
at least 20 bytes, so the next enlargement will probably go in the same link.
If not, there will be another link. You may check how many links there are
with @code{myisamchk -ed}. All links may be removed with @code{myisamchk -r}.
@end itemize


@node Compressed format,  , Dynamic format, MyISAM table formats
@subsubsection Compressed Table Characteristics

@cindex tables, compressed format

This is a read-only type that is generated with the optional
@code{myisampack} tool (@code{pack_isam} for @code{ISAM} tables):

@itemize @bullet
@item
All MySQL distributions, even those that existed before MySQL
went GPL, can read tables that were compressed with @code{myisampack}.
@item
Compressed tables take very little disk space. This minimises disk usage, which
is very nice when using slow disks (like CD-ROMs).
@item
Each record is compressed separately (very little access overhead).  The
header for a record is fixed (1-3 bytes) depending on the biggest record in the
table.  Each column is compressed differently. Some of the compression types
are:
@itemize @minus
@item
There is usually a different Huffman table for each column.
@item
Suffix space compression.
@item
Prefix space compression.
@item
Numbers with value @code{0} are stored using 1 bit.
@item
If values in an integer column have a small range, the column is stored using
the smallest possible type. For example, a @code{BIGINT} column (8 bytes) may
be stored as a @code{TINYINT} column (1 byte) if all values are in the range
@code{0} to @code{255}.
@item
If a column has only a small set of possible values, the column type is
converted to @code{ENUM}.
@item
A column may use a combination of the above compressions.
@end itemize
@item
Can handle fixed- or dynamic-length records.
@item
Can be uncompressed with @code{myisamchk}.
@end itemize


@node MyISAM table problems,  , MyISAM table formats, MyISAM
@subsection MyISAM table problems.

The file format that MySQL uses to store data has been extensively
tested, but there are always circumstances that may cause database tables
to become corrupted.

@menu
* Corrupted MyISAM tables::     Corrupted MyISAM tables.
* MyISAM table close::          Clients is using or hasn't closed the table properly
@end menu


@node Corrupted MyISAM tables, MyISAM table close, MyISAM table problems, MyISAM table problems
@subsubsection Corrupted MyISAM tables.

Even if the MyISAM table format is very reliable (all changes to a table
is written before the SQL statements returns) , you can still get
corrupted tables if some of the following things happens:

@itemize @bullet
@item
The @code{mysqld} process being killed in the middle of a write.
@item
Unexpected shutdown of the computer (for example, if the computer is turned
off).
@item
A hardware error.
@item
You are using an external program (like myisamchk) on a live table.
@item
A software bug in the MySQL or MyISAM code.
@end itemize

Typial typical symptoms for a corrupt table is:

@itemize @bullet
@item
You get the error @code{Incorrect key file for table: '...'. Try to repair it}
while selecting data from the table.
@item
Queries doesn't find rows in the table or returns incomplete data.
@end itemize

You can check if a table is ok with the command @code{CHECK
TABLE}. @xref{CHECK TABLE}.

You can repair a corrupted table with @code{REPAIR TABLE}. @xref{REPAIR TABLE}.
You can also repair a table, when @code{mysqld} is not running with
the @code{myisamchk} command. @code{myisamchk syntax}.

If your tables get corrupted a lot you should try to find the reason for
this! @xref{Crashing}.

In this case the most important thing to know is if the table got
corrupted if the @code{mysqld} died (one can easily verify this by
checking if there is a recent row @code{restarted mysqld} in the mysqld
error file).  If this isn't the case, then you should try to make a test
case of this.  @xref{Reproduceable test case}.


@node MyISAM table close,  , Corrupted MyISAM tables, MyISAM table problems
@subsubsection Clients is using or hasn't closed the table properly

Each @code{MyISAM} @file{.MYI} file has in the header a counter that can
be used to check if a table has been closed properly.

If you get the following warning from @code{CHECK TABLE} or @code{myisamchk}:

@example
# clients is using or hasn't closed the table properly
@end example

this means that this counter has come out of sync.  This doesn't mean
that the table is corrupted, but means that you should at least do a
check on the table to verify that it's okay.

The counter works as follows:

@itemize @bullet
@item
The first time a table is updated in MySQL, a counter in the
header of the index files is incremented.
@item
The counter is not changed during further updates.
@item
When the last instance of a table is closed (because of a @code{FLUSH} or
because there isn't room in the table cache) the counter is
decremented if the table has been updated at any point.
@item
When you repair the table or check the table and it was okay,
the counter is reset to 0.
@item
To avoid problems with interaction with other processes that may do a
check on the table, the counter is not decremented on close if it was 0.
@end itemize

In other words, the only ways this can go out of sync are:

@itemize @bullet
@item
The @code{MyISAM} tables are copied without a @code{LOCK} and
@code{FLUSH TABLES}.
@item
MySQL has crashed between an update and the final close.
(Note that the table may still be okay, as MySQL always issues writes
for everything between each statement.)
@item
Someone has done a @code{myisamchk --repair} or @code{myisamchk
--update-state}on a table that was in use by @code{mysqld}.
@item
Many @code{mysqld} servers are using the table and one has done a
@code{REPAIR} or @code{CHECK} of the table while it was in use by
another server.  In this setup the @code{CHECK} is safe to do (even if
you will get the warning from other servers), but @code{REPAIR} should
be avoided as it currently replaces the datafile with a new one, which
is not signaled to the other servers.
@end itemize


@node MERGE, ISAM, MyISAM, Table types
@section MERGE Tables

@cindex tables, merging
@cindex MERGE tables, defined

@code{MERGE} tables are new in MySQL Version 3.23.25. The code
is still in gamma, but should be resonable stable.

A @code{MERGE} table (also known as a @code{MRG_MyISAM} table) is a
collection of identical @code{MyISAM} tables that can be used as one.
You can only @code{SELECT}, @code{DELETE}, and @code{UPDATE} from the
collection of tables.  If you @code{DROP} the @code{MERGE} table, you
are only dropping the @code{MERGE} specification.

Note that @code{DELETE FROM merge_table} used without a @code{WHERE}
will only clear the mapping for the table, not delete everything in the
mapped tables. (We plan to fix this in 4.1).

With identical tables we mean that all tables are created with identical
column and key information.  You can't merge tables in which the
columns are packed differently, doesn't have exactly the same columns,
or have the keys in different order.  However, some of the tables can be
compressed with @code{myisampack}.  @xref{myisampack, , @code{myisampack}}.

When you create a @code{MERGE} table, you will get a @file{.frm} table
definition file and a @file{.MRG} table list file.  The @file{.MRG} just
contains a list of the index files (@file{.MYI} files) that should
be used as one.  All used tables must be in the same database as the
@code{MERGE} table itself.

For the moment, you need to have @strong{select}, @strong{update}, and
@strong{delete} privileges on the tables you map to a @code{MERGE} table.

@code{MERGE} tables can help you solve the following problems:

@itemize @bullet
@item
Easily manage a set of log tables. For example, you can put data from
different months into separate files, compress some of them with
@code{myisampack}, and then create a @code{MERGE} to use these as one.
@item
Give you more speed. You can split a big read-only table based on some
criteria and then put the different table part on different disks.
A @code{MERGE} table on this could be much faster than using
the big table. (You can, of course, also use a RAID to get the same
kind of benefits.)
@item
Do more efficient searches. If you know exactly what you are looking
after, you can search in just one of the split tables for some queries
and use @code{MERGE} table for others.  You can even have many
different @code{MERGE} tables active, with possible overlapping files.
@item
More efficient repairs. It's easier to repair the individual files that
are mapped to a @code{MERGE} file than trying to repair a real big file.
@item
Instant mapping of many files as one. A @code{MERGE} table uses the
index of the individual tables. It doesn't need to maintain an index of
its one.  This makes @code{MERGE} table collections VERY fast to make or
remap.  Note that you must specify the key definitions when you create
a @code{MERGE} table!.
@item
If you have a set of tables that you join to a big table on demand or
batch, you should instead create a @code{MERGE} table on them on demand.
This is much faster and will save a lot of disk space.
@item
Go around the file-size limit for the operating system.
@item
You can create an alias/synonym for a table by just using @code{MERGE}
over one table. There shouldn't be any really notable performance
impacts of doing this (only a couple of indirect calls and @code{memcpy()}
calls for each read).
@end itemize

The disadvantages with @code{MERGE} tables are:

@itemize @bullet
@item
You can only use identical @code{MyISAM} tables for a @code{MERGE} table.
@item
@code{AUTO_INCREMENT} columns are not automatically updated on @code{INSERT}.
@item
@code{REPLACE} doesn't work.
@item
@code{MERGE} tables uses more file descriptors. If you are using a
@code{MERGE} that maps over 10 tables and 10 users are using this, you
are using 10*10 + 10 file descriptors.  (10 datafiles for 10 users
and 10 shared index files.)
@item
Key reads are slower. When you do a read on a key, the @code{MERGE}
handler will need to issue a read on all underlying tables to check
which one most closely matches the given key.  If you then do a 'read-next'
then the merge table handler will need to search the read buffers
to find the next key. Only when one key buffer is used up, the handler
will need to read the next key block. This makes @code{MERGE} keys much slower
on @code{eq_ref} searches, but not much slower on @code{ref} searches.
@xref{EXPLAIN}.
@item
You can't do @code{DROP TABLE}, @code{ALTER TABLE} or @code{DELETE FROM
table_name} without a @code{WHERE} clause on any of the table that is
mapped by a @code{MERGE} table that is 'open'.  If you do this, the
@code{MERGE} table may still refer to the original table and you will
get unexpected results.
@end itemize

When you create a @code{MERGE} table you have to specify with
@code{UNION(list-of-tables)} which tables you want to use as
one. Optionally you can specify with @code{INSERT_METHOD} if you want
insert for the @code{MERGE} table to happen in the first or last table
in the @code{UNION} list. If you don't specify @code{INSERT_METHOD} or
specify @code{NO}, then all @code{INSERT} commands on the @code{MERGE}
table will return an error.

The following example shows you how to use @code{MERGE} tables:

@example
CREATE TABLE t1 (a INT AUTO_INCREMENT PRIMARY KEY, message CHAR(20));
CREATE TABLE t2 (a INT AUTO_INCREMENT PRIMARY KEY, message CHAR(20));
INSERT INTO t1 (message) VALUES ("Testing"),("table"),("t1");
INSERT INTO t2 (message) VALUES ("Testing"),("table"),("t2");
CREATE TABLE total (a INT NOT NULL, message CHAR(20), KEY(a))
             TYPE=MERGE UNION=(t1,t2) INSERT_METHOD=LAST;
@end example

Note that we didn't create a @code{UNIQUE} or @code{PRIMARY KEY} in the
@code{total} table as the key isn't going to be unique in the @code{total}
table.

Note that you can also manipulate the @file{.MRG} file directly from
the outside of the MySQL server:

@example
shell> cd /mysql-data-directory/current-database
shell> ls -1 t1.MYI t2.MYI > total.MRG
shell> mysqladmin flush-tables
@end example

Now you can do things like:

@example
mysql> SELECT * FROM total;
+---+---------+
| a | message |
+---+---------+
| 1 | Testing |
| 2 | table   |
| 3 | t1      |
| 1 | Testing |
| 2 | table   |
| 3 | t2      |
+---+---------+
@end example

To remap a @code{MERGE} table you can do one of the following:

@itemize @bullet
@item
@code{DROP} the table and re-create it
@item
Use @code{ALTER TABLE table_name UNION(...)}
@item
Change the @file{.MRG} file and issue a @code{FLUSH TABLE} on the
@code{MERGE} table and all underlying tables to force the handler to
read the new definition file.
@end itemize

@menu
* MERGE table problems::        MERGE table problems.
@end menu

@node MERGE table problems,  , MERGE, MERGE
@subsection MERGE table problems.

The following are the known problems with @code{MERGE} tables:

@itemize @bullet
@item
@code{MERGE} table cannot maintain UNIQUE constraints over the whole
table.  When you do INSERT, the data goes into the first or last table
(according to @code{INSERT_METHOD=xxx}) and this MyISAM table ensures
that the data are unique, but it knows nothing about the first MyISAM table.
@item
@code{DELETE FROM merge_table} used without a @code{WHERE}
will only clear the mapping for the table, not delete everything in the
mapped tables.
@item
@code{RENAME TABLE} on a table used in an active @code{MERGE} table may
corrupt the table.  This will be fixed in MySQL 4.0.x.
@item
Creation of a table of type @code{MERGE} doesn't check if the underlying
tables are of compatible types.  If you use @code{MERGE} tables in this
fashion, you are very likely to run into strange problems.
@item
If you use @code{ALTER TABLE} to first add an @code{UNIQUE} index to a
table used in a @code{MERGE} table and then use @code{ALTER TABLE} to
add a normal index on the @code{MERGE} table, the key order will be
different for the tables if there was an old non-unique key in the
table. This is because @code{ALTER TABLE} puts @code{UNIQUE} keys before
normal keys to be able to detect duplicate keys as early as possible.
@item
The range optimizer can't yet use @code{MERGE} table efficiently and may
sometimes produce non-optimal joins.  This will be fixed in MySQL 4.0.x.
@item
@code{DROP TABLE} on a table that is in use by a @code{MERGE} table will
not work on Windows because the @code{MERGE} handler does the table mapping
hidden from the upper layer of MySQL.  Because Windows doesn't allow you
to drop files that are open, you first must flush all @code{MERGE}
tables (with @code{FLUSH TABLES}) or drop the @code{MERGE} table before
dropping the table.  We will fix this at the same time we introduce
@code{VIEW}s.
@end itemize

@node ISAM, HEAP, MERGE, Table types
@section ISAM Tables

@cindex tables, ISAM

You can also use the deprecated @code{ISAM} table type. This will disappear
rather soon (probably in MySQL 4.1) because @code{MyISAM} is a better
implementation of the same thing. @code{ISAM} uses a @code{B-tree} index. The
index is stored in a file with the @file{.ISM} extension, and the data
is stored in a file with the @file{.ISD} extension.  You can
check/repair @code{ISAM} tables with the @code{isamchk} utility. @xref{Crash
recovery}.

@code{ISAM} has the following features/properties:

@itemize @bullet
@item Compressed and fixed-length keys
@item Fixed and dynamic record length
@item 16 keys with 16 key parts/key
@item Max key length 256 (default)
@item Data is stored in machine format; this is fast, but is machine/OS dependent.
@end itemize

Most of the things true for @code{MyISAM} tables are also true for @code{ISAM}
tables. @xref{MyISAM}. The major differences compared to @code{MyISAM}
tables are:

@itemize @bullet
@item @code{ISAM} tables are not binary portable across OS/Platforms.
@item Can't handle tables > 4G.
@item Only support prefix compression on strings.
@item Smaller key limits.
@item Dynamic tables get more fragmented.
@item Tables are compressed with @code{pack_isam} rather than with @code{myisampack}.
@end itemize

If you want to convert an @code{ISAM} table to a @code{MyISAM} table so
that you can use utilities such as @code{mysqlcheck}, use an @code{ALTER
TABLE} statement:

@example
mysql> ALTER TABLE tbl_name TYPE = MYISAM;
@end example

The embedded MySQL versions doesn't support @code{ISAM} tables.

@node HEAP, InnoDB, ISAM, Table types
@section HEAP Tables

@cindex tables, @code{HEAP}

@code{HEAP} tables use a hashed index and are stored in memory.  This
makes them very fast, but if MySQL crashes you will lose all
data stored in them.  @code{HEAP} is very useful for temporary tables!

The MySQL internal @code{HEAP} tables use 100% dynamic hashing
without overflow areas. There is no extra space needed for free lists.
@code{HEAP} tables also don't have problems with delete + inserts, which
normally is common with hashed tables:

@example
mysql> CREATE TABLE test TYPE=HEAP SELECT ip,SUM(downloads) AS down
    ->                   FROM log_table GROUP BY ip;
mysql> SELECT COUNT(ip),AVG(down) FROM test;
mysql> DROP TABLE test;
@end example

Here are some things you should consider when you use @code{HEAP} tables:

@itemize @bullet
@item
You should always use specify @code{MAX_ROWS} in the @code{CREATE} statement
to ensure that you accidentally do not use all memory.
@item
Indexes will only be used with @code{=} and @code{<=>} (but are VERY fast).
@item
@code{HEAP} tables can only use whole keys to search for a row; compare this
to @code{MyISAM} tables where any prefix of the key can be used to find rows.
@item
@code{HEAP} tables use a fixed record length format.
@item
@code{HEAP} doesn't support @code{BLOB}/@code{TEXT} columns.
@item
@code{HEAP} doesn't support @code{AUTO_INCREMENT} columns.
@item
@code{HEAP} doesn't support an index on a @code{NULL} column.
@item
You can have non-unique keys in a @code{HEAP} table (this isn't common for
hashed tables).
@item
@code{HEAP} tables are shared between all clients (just like any other
table).
@item
You can't search for the next entry in order (that is, to use the index
to do an @code{ORDER BY}).
@item
Data for @code{HEAP} tables are allocated in small blocks. The tables
are 100% dynamic (on inserting). No overflow areas and no extra key
space are needed.  Deleted rows are put in a linked list and are
reused when you insert new data into the table.
@item
You need enough extra memory for all HEAP tables that you want to use at
the same time.
@item
To free memory, you should execute @code{DELETE FROM heap_table},
@code{TRUNCATE heap_table} or @code{DROP TABLE heap_table}.
@item
MySQL cannot find out approximately how many rows there
are between two values (this is used by the range optimiser to decide which
index to use).  This may affect some queries if you change a @code{MyISAM}
table to a @code{HEAP} table.
@item
To ensure that you accidentally don't do anything foolish, you can't create
@code{HEAP} tables bigger than @code{max_heap_table_size}.
@end itemize

The memory needed for one row in a @code{HEAP} table is:

@example
SUM_OVER_ALL_KEYS(max_length_of_key + sizeof(char*) * 2)
+ ALIGN(length_of_row+1, sizeof(char*))
@end example

@code{sizeof(char*)} is 4 on 32-bit machines and 8 on 64-bit machines.


@node InnoDB, BDB, HEAP, Table types
@section InnoDB Tables

@menu
* InnoDB overview::             InnoDB Tables Overview
* InnoDB start::                InnoDB Startup Options
* InnoDB init::                 Creating InnoDB Tablespace
* Using InnoDB tables::         Creating InnoDB Tables
* Adding and removing::         Adding and Removing InnoDB Data and Log Files
* Backing up::                  Backing up and Recovering an InnoDB Database
* Moving::                      Moving an InnoDB Database to Another Machine
* InnoDB transaction model::    InnoDB Transaction Model.
* Implementation::              Implementation of Multi-versioning
* Table and index::             Table and Index Structures
* File space management::       File Space Management and Disk I/O
* Error handling::              Error Handling
* InnoDB restrictions::         Restrictions on InnoDB Tables
* InnoDB contact information::  InnoDB Contact Information.
@end menu


@node InnoDB overview, InnoDB start, InnoDB, InnoDB
@subsection InnoDB Tables Overview

@cindex transactions, support
@cindex transaction-safe tables
@cindex ACID

InnoDB provides MySQL with a transaction-safe (@code{ACID} compliant)
table handler with commit, rollback, and crash recovery capabilities.
InnoDB does locking on row level and also provides an Oracle-style
consistent
non-locking read in @code{SELECT}s. These features increase
multiuser concurrency and performance. There is no need for
lock escalation in InnoDB,
because row level locks in InnoDB fit in very small space.
InnoDB tables support @code{FOREIGN KEY} constraints
as the first table type in MySQL.

InnoDB has been designed for maximum performance
when processing
large data volumes. Its CPU efficiency is probably not
matched by any other disk-based relational database engine.

Technically, InnoDB is a complete database backend placed under MySQL.
InnoDB has its own buffer pool for caching data and indexes in main
memory. InnoDB stores its tables and indexes in a tablespace, which
may consist of several files. This is different from, for example,
MyISAM tables where each table is stored as a separate file.
InnoDB tables can be of any size also on those operating
systems where file-size is limited to 2 GB.

You can find the latest information about InnoDB at
@uref{http://www.innodb.com/}. The most up-to-date version of the
InnoDB manual is always placed there, and you can also order
commercial licenses and support for InnoDB.

InnoDB is currently (October 2001) used in production at
several large database sites requiring high performance.
The famous Internet news site Slashdot.org runs on
InnoDB. Mytrix, Inc. stores over 1 TB of data in
InnoDB, and another site handles an average
load of 800 inserts/updates per second in InnoDB.

InnoDB tables are included in the MySQL source distribution
starting from 3.23.34a and are activated in the MySQL -Max
binary. For Windows the -Max binaries are contained in the
standard distribution.

If you have downloaded a binary version of MySQL that includes
support for InnoDB, simply follow the instructions of the
MySQL manual
for installing a binary version of MySQL. If you already have
MySQL-3.23 installed, then the simplest way to install
MySQL -Max is to replace the server executable @file{mysqld}
with the corresponding executable in the -Max distribution.
MySQL and MySQL -Max differ only in the server executable.
@xref{Installing binary}.
@xref{mysqld-max, , @code{mysqld-max}}.

To compile MySQL with InnoDB support,
download MySQL-3.23.34a or newer version from
@uref{http://www.mysql.com/}
and configure MySQL with the
@code{--with-innodb} option. See the
MySQL manual
about installing a MySQL source distribution.
@xref{Installing source}.

@example
cd /path/to/source/of/mysql-3.23.37
./configure --with-innodb
@end example

To use InnoDB you have to specify InnoDB startup options in
your @file{my.cnf} or @file{my.ini} file. The minimal way
to modify it is to add to the @code{[mysqld]} section the line

@example
innodb_data_file_path=ibdata:30M
@end example

but to get good performance it is best that you specify options
as recommended. @xref{InnoDB start}.

InnoDB is distributed under the GNU GPL License Version 2 (of June 1991).
In the source distribution of MySQL, InnoDB appears as a subdirectory.


@node InnoDB start, InnoDB init, InnoDB overview, InnoDB
@subsection InnoDB Startup Options


To use InnoDB tables in MySQL-Max-3.23 you MUST specify configuration
parameters
in the @code{[mysqld]} section of
the configuration file @file{my.cnf}, or on Windows optionally in
@file{my.ini}.

At the minimum, in 3.23 you must specify @code{innodb_data_file_path}.
In MySQL-4.0 you do not need to specify even
@code{innodb_data_file_path}: the default for it is to create
an auto-extending 16 MB file @file{ibdata1} to the @code{datadir}
of MySQL. (In MySQL-4.0.0 and 4.0.1 the datafile is 64 MB and not
auto-extending.)

But to get good performance you MUST explicitly set the InnoDB parameters
listed in the following examples.

Starting from versions 3.23.50 and 4.0.2 InnoDB allows the last
datafile on the @code{innodb_data_file_path} line
to be specified as @strong{auto-extending}. The syntax for
@code{innodb_data_file_path} is then the following:
@example
pathtodatafile:sizespecification;pathtodatafile:sizespecification;...
...  ;pathtodatafile:sizespecification[:autoextend[:max:sizespecification]]
@end example
If you specify the last datafile with the autoextend option, InnoDB
will extend the last datafile if it runs out of free space in the
tablespace. The increment is 8 MB at a time. An example:
@example
innodb_data_file_path = /ibdata/ibdata1:100M:autoextend
@end example
instructs InnoDB to create just a single datafile whose initial size is
100 MB and which is extended in 8 MB blocks when space runs out.
If the disk becomes full you may want to add another data
file to another disk, for example. Then you have to look the size
of @file{ibdata1}, round the size downward to
the closest multiple of 1024 * 1024 bytes (= 1 MB), and specify
the rounded size of @file{ibdata1} explicitly in
@code{innodb_data_file_path}.
After that you can add another datafile:
@example
innodb_data_file_path = /ibdata/ibdata1:988M;/disk2/ibdata2:50M:autoextend
@end example
Be cautious on filesystems where the maximum file-size is 2 GB!
InnoDB is not aware of the OS maximum file-size. On those filesystems
you might want to specify the max size for the datafile:
@example
innodb_data_file_path = /ibdata/ibdata1:100M:autoextend:max:2000M
@end example

Suppose you have a Windows NT computer with 128 MB RAM and a
single 10 GB hard disk.
Following is an example of possible configuration parameters in
@file{my.cnf} or @file{my.ini} for InnoDB:

@example
[mysqld]
# You can write your other MySQL server options here
# ...
#
innodb_data_home_dir = c:\ibdata
#                                  Datafiles must be able to
#                                  hold your data and indexes
innodb_data_file_path = ibdata1:2000M;ibdata2:2000M
#                                  Set buffer pool size to 50 - 80 %
#                                  of your computer's memory
set-variable = innodb_buffer_pool_size=70M
set-variable = innodb_additional_mem_pool_size=10M
innodb_log_group_home_dir = c:\iblogs
#                                  .._log_arch_dir must be the same
#                                  as .._log_group_home_dir
innodb_log_arch_dir = c:\iblogs
innodb_log_archive=0
set-variable = innodb_log_files_in_group=3
#                                  Set the log file-size to about
#                                  15 % of the buffer pool size
set-variable = innodb_log_file_size=10M
set-variable = innodb_log_buffer_size=8M
#                                  Set ..flush_log_at_trx_commit to
#                                  0 if you can afford losing
#                                  a few last transactions 
innodb_flush_log_at_trx_commit=1
set-variable = innodb_file_io_threads=4
set-variable = innodb_lock_wait_timeout=50
@end example

Note that @code{InnoDB} @strong{does not create directories:
you must create them yourself.}
Use the Unix or MS-DOS @code{mkdir} command to create
the data and log group home directories.
Check also that the MySQL server
has @strong{the rights to create files} in the directories you specify.

Note that datafiles must be < 2G in
some filesystems! The combined size of datafiles
must be >= 10 MB.
The combined size of the log files must be < 4G.

If you do not specify @code{innodb_data_home_dir}, then
the default is that InnoDB creates its datafiles to the
@code{datadir} of MySQL. Then you cannot use absolute
file paths in @code{innodb_data_file_path}.

When you the first time create an InnoDB database, it
is best that you start the MySQL server from the command
prompt. Then InnoDB will print the information about the
database creation to the screen, and you see what is
happening.
For example, in Windows you can start @file{mysqld-max.exe} with:
@example
your-path-to-mysqld>mysqld-max --standalone --console
@end example
For information about what the printout should look like,
see @ref{InnoDB init}.

@strong{Where to put @file{my.cnf} or @file{my.ini} in Windows?}
The rules for Windows are the following:
@itemize bullet
@item Only one of @file{my.cnf} or @file{my.ini} should be created.
@item The @file{my.cnf} file should be placed in the root
directory of the drive @file{C:}.
@item The @file{my.ini} file should be placed in the WINDIR directory, e.g,
@file{C:\WINDOWS} or @file{C:\WINNT}. You can use the @code{SET}
command of MS-DOS to print the value of WINDIR.
@item  If your PC uses a boot loader where the @file{C:} drive
is not the boot drive, then your only option is to use the @file{my.ini} file.
@end itemize

@strong{Where to specify options in Unix?}
On Unix @file{mysqld} reads options from the following files, if they exist,
in the following order: 
@itemize bullet
@item @file{/etc/my.cnf}  Global options.  
@item @file{COMPILATION_DATADIR/my.cnf}  Server-specific options.  
@item @file{defaults-extra-file}  The file specified with
@code{--defaults-extra-file=...}.
@item @file{~/.my.cnf} User-specific options.
@end itemize
@file{COMPILATION_DATADIR} is the MySQL data directory which was
specified as a @code{./configure} option when @file{mysqld}
was compiled
(typically @file{/usr/local/mysql/data} for a binary installation or @file{/usr/local/var} for a source installation).

If you are not sure from where @file{mysqld} reads its @file{my.cnf}
or @file{my.ini}, you can give the path as the first command-line
option to the server:
@code{mysqld --defaults-file=your_path_to_my_cnf}.

Suppose you have a Linux computer with 512 MB RAM and
three 20 GB hard disks (at directory paths @file{`/'},
@file{`/dr2'} and @file{`/dr3'}).
Here is an example of possible configuration parameters
in @file{my.cnf} for
InnoDB:

@example
[mysqld]
# You can write your other MySQL server options here
# ...
#
innodb_data_home_dir = /
#                                  Datafiles must be able to
#                                  hold your data and indexes
innodb_data_file_path = ibdata/ibdata1:2000M;dr2/ibdata/ibdata2:2000M
#                                  Set buffer pool size to 50 - 80 %
#                                  of your computer's memory, but
#                                  make sure on Linux x86 total
#                                  memory usage is < 2 GB
set-variable = innodb_buffer_pool_size=350M
set-variable = innodb_additional_mem_pool_size=20M
innodb_log_group_home_dir = /dr3/iblogs
#                                  .._log_arch_dir must be the same
#                                  as .._log_group_home_dir
innodb_log_arch_dir = /dr3/iblogs
innodb_log_archive=0
set-variable = innodb_log_files_in_group=3
#                                  Set the log file-size to about
#                                  15 % of the buffer pool size
set-variable = innodb_log_file_size=50M
set-variable = innodb_log_buffer_size=8M
#                                  Set ..flush_log_at_trx_commit to
#                                  0 if you can afford losing
#                                  a few last transactions 
innodb_flush_log_at_trx_commit=1
set-variable = innodb_file_io_threads=4
set-variable = innodb_lock_wait_timeout=50
#innodb_flush_method=fdatasync
#innodb_fast_shutdown=1
#set-variable = innodb_thread_concurrency=5
@end example

Note that we have placed the two datafiles on different disks.
The reason for the name @code{innodb_data_file_path} is that
you can also specify paths to your datafiles, and
@code{innodb_data_home_dir} is just textually catenated
before your datafile paths, adding a possible slash or
backslash in between. InnoDB will fill the tablespace
formed by the datafiles from bottom up. In some cases it will
improve the performance of the database if all data is not placed
on the same physical disk. Putting log files on a different disk from
data is very often beneficial for performance.
You can also use @strong{raw disk partitions} (raw devices)
as datafiles. In some Unixes
they speed up I/O. See the manual section on InnoDB file space management 
about how to specify them in @file{my.cnf}.

@strong{Warning}: on Linux x86 you must be careful you @strong{do not set memory usage
too high}. glibc will allow the process heap to grow over thread stacks,
which will crash your server. It is a risk if the value of
@example
innodb_buffer_pool_size + key_buffer +
max_connections * (sort_buffer + record_buffer) + max_connections * 2 MB
@end example
is close to 2 GB or exceeds 2 GB. Each thread will use a stack
(often 2 MB, but in MySQL AB binaries only 256 kB) and in the worst case also
@code{sort_buffer + record_buffer}
additional memory.

@strong{How to tune other @file{mysqld} server parameters?}
Typical values which suit most users are:
@example
set-variable = max_connections=200
set-variable = record_buffer=1M
set-variable = sort_buffer=1M
#                                  Set key_buffer to 5 - 50 %
#                                  of your RAM depending on how
#                                  much you use MyISAM tables, but
#                                  keep key_buffer + InnoDB
#                                  buffer pool size < 80 % of
#                                  your RAM
set-variable = key_buffer=...
@end example

Note that some parameters are given using the numeric @file{my.cnf}
parameter format: @code{set-variable = innodb... = 123}, others
(string and boolean parameters) with another format:
@code{innodb_... = ... }.


The meanings of the configuration parameters are the following:

@multitable @columnfractions .35 .65
@item @strong{Option} @tab @strong{Description}
@item @code{innodb_data_home_dir} @tab
The common part of the directory path for all InnoDB datafiles.
The default for this parameter is the @code{datadir} of MySQL.
@item @code{innodb_data_file_path} @tab
Paths to individual datafiles and their sizes. The full directory path
to each datafile is acquired by concatenating innodb_data_home_dir to
the paths specified here. The file-sizes are specified in megabytes,
hence the 'M' after the size specification above.
InnoDB also understands the abbreviation 'G', 1G meaning 1024M.
Starting from
3.23.44 you can set the file-size bigger than 4 GB on those
operating systems which support big files.
On some operating systems files must be < 2 GB.
The sum of the sizes of the files must be at least 10 MB.
@item @code{innodb_mirrored_log_groups} @tab
Number of identical copies of log groups we
keep for the database. Currently this should be set to 1.
@item @code{innodb_log_group_home_dir} @tab
Directory path to InnoDB log files.
@item @code{innodb_log_files_in_group} @tab
Number of log files in the log group. InnoDB writes to the files in a
circular fashion. Value 3 is recommended here.
@item @code{innodb_log_file_size} @tab
Size of each log file in a log group in megabytes. Sensible values range
from 1M to 1/nth of the size of the buffer pool specified below,
where n is the number of log files in the group. The
bigger the value,
the less checkpoint flush activity is needed in the buffer pool,
saving disk I/O. But bigger log files also mean that recovery will be
slower in case of a crash. The combined size of log files must
be < 4 GB on 32-bit computers.
@item @code{innodb_log_buffer_size} @tab
The size of the buffer which InnoDB uses to write log to the log files
on disk.  Sensible values range from 1M to half the combined size of log
files. A big log buffer allows large transactions to run without a need
to write the log to disk until the transaction commit. Thus, if you have
big transactions, making the log buffer big will save disk I/O.
@item @code{innodb_flush_log_at_trx_commit} @tab
Normally this is set to 1, meaning that at a transaction commit the log
is flushed to disk, and the modifications made by the transaction become
permanent, and survive a database crash. If you are willing to
compromise this safety, and you are running small transactions, you may
set this to 0 to reduce disk I/O to the logs.
@item @code{innodb_log_arch_dir} @tab
The directory where fully written log files would be archived if we used
log archiving.  The value of this parameter should currently be set the
same as @code{innodb_log_group_home_dir}.
@item @code{innodb_log_archive} @tab
This value should currently be set to 0.  As recovery from a backup is
done by MySQL using its own log files, there is currently no need to
archive InnoDB log files.
@item @code{innodb_buffer_pool_size} @tab
The size of the memory buffer InnoDB uses to cache data and indexes of
its tables.  The bigger you set this the less disk I/O is needed to
access data in tables. On a dedicated database server you may set this
parameter up to 80 % of the machine physical memory size. Do not set it
too large, though, because competition of the physical memory may cause
paging in the operating system.
@item @code{innodb_additional_mem_pool_size} @tab
Size of a memory pool InnoDB uses to store data dictionary information
and other internal data structures. A sensible value for this might be
2M, but the more tables you have in your application the more you will
need to allocate here. If InnoDB runs out of memory in this pool, it
will start to allocate memory from the operating system, and write
warning messages to the MySQL error log.
@item @code{innodb_file_io_threads} @tab
Number of file I/O threads in InnoDB. Normally, this should be 4, but
on Windows disk I/O may benefit from a larger number.
@item @code{innodb_lock_wait_timeout} @tab
Timeout in seconds an InnoDB transaction may wait for a lock before
being rolled back.  InnoDB automatically detects transaction deadlocks
in its own lock table and rolls back the transaction. If you use
@code{LOCK TABLES} command, or other transaction-safe table handlers
than InnoDB in the same transaction, then a deadlock may arise which
InnoDB cannot notice. In cases like this the timeout is useful to
resolve the situation.
@item @code{innodb_flush_method} @tab
(Available from 3.23.40 up.)
The default value for this is @code{fdatasync}.
Another option is @code{O_DSYNC}.
@end multitable


@node InnoDB init, Using InnoDB tables, InnoDB start, InnoDB
@subsection Creating InnoDB Tablespace

Suppose you have installed MySQL and have edited @file{my.cnf} so that
it contains the necessary InnoDB configuration parameters.
Before starting MySQL you should check that the directories you have
specified for InnoDB datafiles and log files exist and that you have
access rights to those directories. InnoDB
cannot create directories, only files. Check also you have enough disk space
for the data and log files.

When you now start MySQL, InnoDB will start creating your datafiles
and log files. InnoDB will print something like the following:

@example
~/mysqlm/sql > mysqld
InnoDB: The first specified datafile /home/heikki/data/ibdata1
did not exist:
InnoDB: a new database to be created!
InnoDB: Setting file /home/heikki/data/ibdata1 size to 134217728
InnoDB: Database physically writes the file full: wait...
InnoDB: datafile /home/heikki/data/ibdata2 did not exist:
new to be created
InnoDB: Setting file /home/heikki/data/ibdata2 size to 262144000
InnoDB: Database physically writes the file full: wait...
InnoDB: Log file /home/heikki/data/logs/ib_logfile0 did not exist:
new to be created
InnoDB: Setting log file /home/heikki/data/logs/ib_logfile0 size to 5242880
InnoDB: Log file /home/heikki/data/logs/ib_logfile1 did not exist:
new to be created
InnoDB: Setting log file /home/heikki/data/logs/ib_logfile1 size to 5242880
InnoDB: Log file /home/heikki/data/logs/ib_logfile2 did not exist:
new to be created
InnoDB: Setting log file /home/heikki/data/logs/ib_logfile2 size to 5242880
InnoDB: Started
mysqld: ready for connections
@end example

A new InnoDB database has now been created. You can connect to the MySQL
server with the usual MySQL client programs like @code{mysql}.
When you shut down the MySQL server with @file{mysqladmin shutdown},
InnoDB output will be like the following:

@example
010321 18:33:34  mysqld: Normal shutdown
010321 18:33:34  mysqld: Shutdown Complete
InnoDB: Starting shutdown...
InnoDB: Shutdown completed
@end example

You can now look at the datafiles and logs directories and you
will see the files created. The log directory will also contain
a small file named @file{ib_arch_log_0000000000}. That file
resulted from the database creation, after which InnoDB switched off
log archiving.
When MySQL is again started, the output will be like the following:

@example
~/mysqlm/sql > mysqld
InnoDB: Started
mysqld: ready for connections
@end example

@menu
* Error creating InnoDB::       If something goes wrong in database creation
@end menu


@node Error creating InnoDB,  , InnoDB init, InnoDB init
@subsubsection If Something Goes Wrong in Database Creation

If InnoDB prints an operating system error in a file operation,
usually the problem is one of the following:

@itemize @bullet
@item You did not create InnoDB data or log directories.
@item @file{mysqld} does not have the rights to create files in those
directories.
@item @file{mysqld} does not read the right @file{my.cnf} or @file{my.ini}
file, and consequently does not see the options you specified.
@item The disk is full or a disk quota is exceeded.
@item You have created a subdirectory whose name is equal to a datafile
you specified.
@item There is a syntax error in @code{innodb_data_home_dir}
or @code{innodb_data_file_path}.
@end itemize

If something goes wrong in an InnoDB database creation, you should
delete all files created by InnoDB. This means all datafiles, all log
files, the small archived log file, and in the case you already did
create some InnoDB tables, delete also the corresponding @file{.frm}
files for these tables from the MySQL database
directories. Then you can try the InnoDB database creation again.


@node Using InnoDB tables, Adding and removing, InnoDB init, InnoDB
@subsection Creating InnoDB Tables

Suppose you have started the MySQL client with the command
@code{mysql test}.
To create a table in the InnoDB format you must specify
@code{TYPE = InnoDB} in the table creation SQL command:

@example
CREATE TABLE CUSTOMER (A INT, B CHAR (20), INDEX (A)) TYPE = InnoDB;
@end example

This SQL command will create a table and an index on column @code{A}
into the InnoDB tablespace consisting of the datafiles you specified
in @file{my.cnf}. In addition MySQL will create a file
@file{CUSTOMER.frm} to the MySQL database directory @file{test}.
Internally, InnoDB will add to its own data dictionary an entry
for table @code{'test/CUSTOMER'}. Thus you can create a table
of the same name @code{CUSTOMER} in another database of MySQL, and
the table names will not collide inside InnoDB.

You can query the amount of free space in the InnoDB tablespace
by issuing the table status command of MySQL for any table you have
created with @code{TYPE = InnoDB}. Then the amount of free
space in the tablespace appears in the table comment section in the
output of @code{SHOW}. An example:

@example
SHOW TABLE STATUS FROM test LIKE 'CUSTOMER'
@end example

Note that the statistics @code{SHOW} gives about InnoDB tables
are only approximate: they are used in SQL optimisation. Table and
index reserved sizes in bytes are accurate, though.

@subsubsection Converting MyISAM Tables to InnoDB

InnoDB does not have a special optimisation for separate index creation.
Therefore it does not pay to export and import the table and create indexes
afterwards.
The fastest way to alter a table to InnoDB is to do the inserts
directly to an InnoDB table, that is, use @code{ALTER TABLE ... TYPE=INNODB},
or create an empty InnoDB table with identical definitions and insert
the rows with @code{INSERT INTO ... SELECT * FROM ...}.

To get better control over the insertion process, it may be good to insert
big tables in pieces:

@example
INSERT INTO newtable SELECT * FROM oldtable
   WHERE yourkey > something AND yourkey <= somethingelse;
@end example

After all data has been inserted you can rename the tables.

During the conversion of big tables you should set the InnoDB
buffer pool size big
to reduce disk I/O. Not bigger than 80 % of the physical memory, though.
You should set InnoDB log files big, and also the log buffer large.

Make sure you do not run out of tablespace: InnoDB tables take a lot
more space than MyISAM tables. If an @code{ALTER TABLE} runs out
of space, it will start a rollback, and that can take hours if it is
disk-bound.
In inserts InnoDB uses the insert buffer to merge secondary index records
to indexes in batches. That saves a lot of disk I/O. In rollback no such
mechanism is used, and the rollback can take 30 times longer than the
insertion.

In the case of a runaway rollback, if you do not have valuable data in your
database,
it is better that you kill the database process and delete all InnoDB data
and log files and all InnoDB table @file{.frm} files, and start
your job again, rather than wait for millions of disk I/Os to complete.

@subsubsection  Foreign Key Constraints

Starting from version 3.23.43b InnoDB features foreign key constraints.
InnoDB is the first MySQL table type which allows you to define foreign key
constraints to guard the integrity of your data.

The syntax of a foreign key constraint definition in InnoDB:
@example
FOREIGN KEY (index_col_name, ...)
                  REFERENCES table_name (index_col_name, ...)
                  [ON DELETE CASCADE | ON DELETE SET NULL]
@end example
Both tables have to be InnoDB type and @strong{there must be an index
where the foreign key and the referenced key are listed as the first
columns}. InnoDB does not auto-create indexes on foreign keys or
referenced keys: you have to create them explicitly.

Corresponding columns in the foreign key
and the referenced key must have similar internal data types
inside InnoDB so that they can be compared without a type
conversion.
The size and the signedness of integer types has to be the same.
The length of string types need not be the same.

Starting from version 3.23.50 you can also associate the
@code{ON DELETE CASCADE} or @code{ON DELETE SET NULL}
clause with the foreign key constraint.

If @code{ON DELETE CASCADE} is specified, and a row in the parent
table is deleted, then InnoDB automatically deletes also all those rows
in the child table whose foreign key values are equal to
the referenced key value in the parent row. If @code{ON DELETE SET NULL}
is specified, the child rows are automatically updated so that the
columns in the foreign key are set to the SQL NULL value.

Starting from version 3.23.50, InnoDB does not check foreign key
constraints on those foreign key or referenced key values
which contain a NULL column.

Starting from version 3.23.50 the InnoDB parser allows you to
use backquotes (`) around table and column names in the above
definition but the InnoDB parser is not yet aware of possible
variable @code{lower_case_table_names} you give in @file{my.cnf}.

An example:
@example
CREATE TABLE parent(id INT NOT NULL, PRIMARY KEY (id)) TYPE=INNODB;
CREATE TABLE child(id INT, parent_id INT, INDEX par_ind (parent_id),
             FOREIGN KEY (parent_id) REFERENCES parent(id)
             ON DELETE SET NULL
) TYPE=INNODB;
@end example

If MySQL gives the error number 1005 from a @code{CREATE TABLE}
statement, and the error message string refers to errno 150, then
the table creation failed because a foreign key constraint was not
correctly formed.
Similarly, if an @code{ALTER TABLE} fails and it refers to errno
150, that means a foreign key definition would be incorrectly
formed for the altered table.

Starting from version 3.23.50 InnoDB allows you to add a new
foreign key constraint to a table through
@example
ALTER TABLE yourtablename
  ADD CONSTRAINT FOREIGN KEY (...) REFERENCES anothertablename(...)
@end example
Remember to create the required indexes first, though.
                           
In InnoDB versions < 3.23.50 @code{ALTER TABLE}
or @code{CREATE INDEX}
should not be used in connection with tables which have foreign
key constraints or which are referenced in foreign key constraints:
Any @code{ALTER TABLE} removes all foreign key
constrainst defined for the table. You should not use
@code{ALTER TABLE} to the referenced table either, but
use @code{DROP TABLE} and @code{CREATE TABLE} to modify the
schema. When MySQL does an @code{ALTER TABLE} it may internally
use @code{RENAME TABLE}, and that will confuse the
foreign key costraints which refer to the table.
A @code{CREATE INDEX} statement is in MySQL
processed as an @code{ALTER TABLE}, and these
restrictions apply also to it.

When doing foreign key checks InnoDB sets shared row
level locks on child or parent records it has to look at.
InnoDB checks foreign key constraints immediately: the check
is not deferred to transaction commit.

InnoDB allows you to drop any table even though that
would break the foreign key constraints which reference
the table. When you drop a table the constraints which
were defined in its create statement are also dropped.

If you re-create a table which was dropped, it has to have
a definition which conforms to the foreign key constraints
referencing it. It must have the right column names and types,
and it must have indexes on the referenced keys, as stated above.
If these are not satisfied, MySQL returns error number 1005
and refers to errno 150 in the error message string.

Starting from version 3.23.50 InnoDB returns the foreign key
definitions of a table when you call
@example
SHOW CREATE TABLE yourtablename
@end example
Then also @file{mysqldump} produces correct definitions
of tables to the dump file, and does not forget about the
foreign keys.

You can also list the foreign key constraints for a table
@code{T} with
@example
SHOW TABLE STATUS FROM yourdatabasename LIKE 'T'
@end example
The foreign key constraints are listed in the table comment of
the output.

InnoDB does not yet support @code{ON DELETE CASCADE}
or other special options on the constraints.

@node Adding and removing, Backing up, Using InnoDB tables, InnoDB
@subsection Adding and Removing InnoDB Data and Log Files

You cannot increase the size of an InnoDB datafile. To add more into
your tablespace you have to add a new datafile. To do this you have to
shut down your MySQL database, edit the @file{my.cnf} file, adding a
new file to @code{innodb_data_file_path}, and then start MySQL
again.

Currently you cannot remove a datafile from InnoDB. To decrease the
size of your database you have to use @file{mysqldump} to dump
all your tables, create a new database, and import your tables to the
new database.

If you want to change the number or the size of your InnoDB log files,
you have to shut down MySQL and make sure that it shuts down without errors.
Then copy the old log files into a safe place just in case something
went wrong in the shutdown and you will need them to recover the
database. Delete then the old log files from the log file directory,
edit @file{my.cnf}, and start MySQL again. InnoDB will tell
you at the startup that it is creating new log files.


@node Backing up, Moving, Adding and removing, InnoDB
@subsection Backing up and Recovering an InnoDB Database

The key to safe database management is taking regular backups.

InnoDB Hot Backup is an online backup tool you can use to
backup your InnoDB database while it is running. InnoDB
Hot Backup does not require you to shut down your database
and it does not set any locks or disturb your normal
database processing. InnoDB Hot Backup is a non-free
additional tool which is not included in the standard
MySQL distribution. See the InnoDB Hot Backup homepage
@uref{http://www.innodb.com/hotbackup.html}
for detailed information and screenshots. 

If you are able to shut down your MySQL server, then to take
a 'binary' backup of your database you have to do the following: 

@itemize @bullet
@item
Shut down your MySQL database and make sure it shuts down without errors.
@item
Copy all your datafiles into a safe place.
@item
Copy all your InnoDB log files to a safe place.
@item
Copy your @file{my.cnf} configuration file(s) to a safe place.
@item
Copy all the @file{.frm} files for your InnoDB tables into a
safe place.
@end itemize

There is currently no online or incremental backup tool available for
InnoDB, though they are in the TODO list.

In addition to taking the binary backups described above,
you should also regularly take dumps of your tables with
@file{mysqldump}. The reason to this is that a binary file
may be corrupted without you noticing it. Dumped tables are stored
into text files which are human-readable and much simpler than
database binary files. Seeing table corruption from dumped files
is easier, and since their format is simpler, the chance for
serious data corruption in them is smaller.

A good idea is to take the dumps at the same time you take a binary
backup of your database. You have to shut out all clients from your
database to get a consistent snapshot of all your tables into your
dumps. Then you can take the binary backup, and you will then have
a consistent snapshot of your database in two formats.

To be able to recover your InnoDB database to the present from the
binary backup described above, you have to run your MySQL database
with the general logging and log archiving of MySQL switched on. Here
by the general logging we mean the logging mechanism of the MySQL server
which is independent of InnoDB logs.

To recover from a crash of your MySQL server process, the only thing
you have to do is to restart it. InnoDB will automatically check the
logs and perform a roll-forward of the database to the present.
InnoDB will automatically roll back uncommitted transactions which were
present at the time of the crash. During recovery, InnoDB will print
out something like the following:

@example
~/mysqlm/sql > mysqld
InnoDB: Database was not shut down normally.
InnoDB: Starting recovery from log files...
InnoDB: Starting log scan based on checkpoint at
InnoDB: log sequence number 0 13674004
InnoDB: Doing recovery: scanned up to log sequence number 0 13739520
InnoDB: Doing recovery: scanned up to log sequence number 0 13805056
InnoDB: Doing recovery: scanned up to log sequence number 0 13870592
InnoDB: Doing recovery: scanned up to log sequence number 0 13936128
...
InnoDB: Doing recovery: scanned up to log sequence number 0 20555264
InnoDB: Doing recovery: scanned up to log sequence number 0 20620800
InnoDB: Doing recovery: scanned up to log sequence number 0 20664692
InnoDB: 1 uncommitted transaction(s) which must be rolled back
InnoDB: Starting rollback of uncommitted transactions
InnoDB: Rolling back trx no 16745
InnoDB: Rolling back of trx no 16745 completed
InnoDB: Rollback of uncommitted transactions completed
InnoDB: Starting an apply batch of log records to the database...
InnoDB: Apply batch completed
InnoDB: Started
mysqld: ready for connections
@end example

If your database gets corrupted or your disk fails, you have
to do the recovery from a backup. In the case of corruption, you should
first find a backup which is not corrupted. From a backup do the recovery
from the general log files of MySQL according to instructions in the
MySQL manual.

@menu
* InnoDB checkpoints::          Checkpoints
@end menu


@node InnoDB checkpoints,  , Backing up, Backing up
@subsubsection Checkpoints

InnoDB implements a checkpoint mechanism called a fuzzy
checkpoint. InnoDB will flush modified database pages from the buffer
pool in small batches, there is no need to flush the buffer pool
in one single batch, which would in practice stop processing
of user SQL statements for a while.

In crash recovery InnoDB looks for a checkpoint label written
to the log files. It knows that all modifications to the database
before the label are already present on the disk image of the database.
Then InnoDB scans the log files forward from the place of the checkpoint
applying the logged modifications to the database.

InnoDB writes to the log files in a circular fashion.
All committed modifications which make the database pages in the buffer
pool different from the images on disk must be available in the log files
in case InnoDB has to do a recovery. This means that when InnoDB starts
to reuse a log file in the circular fashion, it has to make sure that the
database page images on disk already contain the modifications
logged in the log file InnoDB is going to reuse. In other words, InnoDB
has to make a checkpoint and often this involves flushing of
modified database pages to disk.

The above explains why making your log files very big may save
disk I/O in checkpointing. It can make sense to set
the total size of the log files as big as the buffer pool or even bigger.
The drawback in big log files is that crash recovery can last longer
because there will be more log to apply to the database.


@node Moving, InnoDB transaction model, Backing up, InnoDB
@subsection Moving an InnoDB Database to Another Machine

InnoDB data and log files are binary-compatible on all platforms
if the floating-point number format on the machines is the same.
You can move an InnoDB database simply by copying all the relevant
files, which we already listed in the previous section on backing up
a database. If the floating-point formats on the machines are
different but you have not used @code{FLOAT} or @code{DOUBLE}
data types in your tables then the procedure is the same: just copy
the relevant files. If the formats are different and your tables
contain floating-point data, you have to use @file{mysqldump}
and @file{mysqlimport} to move those tables.

A performance tip is to switch off the auto commit when you import
data into your database, assuming your tablespace has enough space for
the big rollback segment the big import transaction will generate.
Do the commit only after importing a whole table or a segment of
a table.


@node InnoDB transaction model, Implementation, Moving, InnoDB
@subsection InnoDB Transaction Model

In the InnoDB transaction model the goal has been to combine the best
properties of a multi-versioning database to traditional two-phase locking.
InnoDB does locking on row level and runs queries by default
as non-locking consistent reads, in the style of Oracle.
The lock table in InnoDB is stored so space-efficiently that lock
escalation is not needed: typically several users are allowed
to lock every row in the database, or any random subset of the rows,
without InnoDB running out of memory.

In InnoDB all user activity happens inside transactions. If the
auto-commit mode is used in MySQL, then each SQL statement
will form a single transaction. If the auto commit mode is
switched off, then we can think that a user always has a transaction
open. If he issues
the SQL @code{COMMIT} or @code{ROLLBACK} statement, that
ends the current transaction, and a new starts. Both statements
will release all InnoDB locks that were set during the
current transaction. A @code{COMMIT} means that the
changes made in the current transaction are made permanent
and become visible to other users. A @code{ROLLBACK}
on the other hand cancels all modifications made by the current
transaction.

@menu
* InnoDB consistent read::      Consistent Read
* InnoDB locking reads::        Locking Reads
* InnoDB Next-key locking::     Next-key Locking: Avoiding the Phantom Problem
* InnoDB Locks set::            Locks Set by Different SQL Statements in InnoDB
* InnoDB Deadlock detection::   Deadlock Detection and Rollback
* InnoDB Consistent read example::  An Example of How the Consistent Read Works in InnoDB
@end menu


@node InnoDB consistent read, InnoDB locking reads, InnoDB transaction model, InnoDB transaction model
@subsubsection Consistent Read

A consistent read means that InnoDB uses its multi-versioning to
present to a query a snapshot of the database at a point in time.
The query will see the changes made by exactly those transactions that
committed before that point of time, and no changes made by later
or uncommitted transactions. The exception to this rule is that the
query will see the changes made by the transaction itself which issues
the query.

When a transaction issues its first consistent read, InnoDB assigns
the snapshot, or the point of time, which all consistent reads in the
same transaction will use. In the snapshot are all transactions that
committed before assigning the snapshot. Thus the consistent reads
within the same transaction will also be consistent with respect to each
other. You can get a fresher snapshot for your queries by committing
the current transaction and after that issuing new queries.

Consistent read is the default mode in which InnoDB processes
@code{SELECT} statements. A consistent read does not set any locks
on the tables it accesses, and therefore other users are free to
modify those tables at the same time a consistent read is being performed
on the table.


@node InnoDB locking reads, InnoDB Next-key locking, InnoDB consistent read, InnoDB transaction model
@subsubsection Locking Reads

A consistent read is not convenient in some circumstances.
Suppose you want to add a new row into your table @code{CHILD},
and make sure that the child already has a parent in table
@code{PARENT}.

Suppose you use a consistent read to read the table @code{PARENT}
and indeed see the parent of the child in the table. Can you now safely
add the child row to table @code{CHILD}? No, because it may
happen that meanwhile some other user has deleted the parent row
from the table @code{PARENT}, and you are not aware of that.

The solution is to perform the @code{SELECT} in a locking
mode, @code{LOCK IN SHARE MODE}.

@example
SELECT * FROM PARENT WHERE NAME = 'Jones' LOCK IN SHARE MODE;
@end example

Performing a read in share mode means that we read the latest
available data, and set a shared mode lock on the rows we read.
If the latest data belongs to a yet uncommitted transaction of another
user, we will wait until that transaction commits.
A shared mode lock prevents others from updating or deleting
the row we have read. After we see that the above query returns
the parent @code{'Jones'}, we can safely add his child
to table @code{CHILD}, and commit our transaction.
This example shows how to implement referential
integrity in your application code.

Let us look at another example: we have an integer counter field in
a table @code{CHILD_CODES} which we use to assign
a unique identifier to each child we add to table @code{CHILD}.
Obviously, using a consistent read or a shared mode read
to read the present value of the counter is not a good idea, since
then two users of the database may see the same value for the
counter, and we will get a duplicate key error when we add
the two children with the same identifier to the table.

In this case there are two good ways to implement the
reading and incrementing of the counter: (1) update the counter
first by incrementing it by 1 and only after that read it,
or (2) read the counter first with
a lock mode @code{FOR UPDATE}, and increment after that:

@example
SELECT COUNTER_FIELD FROM CHILD_CODES FOR UPDATE;
UPDATE CHILD_CODES SET COUNTER_FIELD = COUNTER_FIELD + 1;
@end example

A @code{SELECT ... FOR UPDATE} will read the latest
available data setting exclusive locks on each row it reads.
Thus it sets the same locks a searched SQL @code{UPDATE} would set
on the rows.


@node InnoDB Next-key locking, InnoDB Locks set, InnoDB locking reads, InnoDB transaction model
@subsubsection Next-key Locking: Avoiding the Phantom Problem

In row level locking InnoDB uses an algorithm called next-key locking.
InnoDB does the row level locking so that when it searches or
scans an index of a table, it sets shared or exclusive locks
on the index records in encounters. Thus the row level locks are
more precisely called index record locks.

The locks InnoDB sets on index records also affect the 'gap'
before that index record. If a user has a shared or exclusive
lock on record R in an index, then another user cannot insert
a new index record immediately before R in the index order.
This locking of gaps is done to prevent the so-called phantom
problem. Suppose I want to read and lock all children with identifier
bigger than 100 from table @code{CHILD},
and update some field in the selected rows.

@example
SELECT * FROM CHILD WHERE ID > 100 FOR UPDATE;
@end example

Suppose there is an index on table @code{CHILD} on column
@code{ID}. Our query will scan that index starting from
the first record where @code{ID} is bigger than 100.
Now, if the locks set on the index records would not lock out
inserts made in the gaps, a new child might meanwhile be
inserted to the table. If now I in my transaction execute

@example
SELECT * FROM CHILD WHERE ID > 100 FOR UPDATE;
@end example

again, I will see a new child in the result set the query returns.
This is against the isolation principle of transactions:
a transaction should be able to run so that the data
it has read does not change during the transaction. If we regard
a set of rows as a data item, then the new 'phantom' child would break
this isolation principle.

When InnoDB scans an index it can also lock the gap
after the last record in the index. Just that happens in the previous
example: the locks set by InnoDB will prevent any insert to
the table where @code{ID} would be bigger than 100.

You can use next-key locking to implement a uniqueness
check in your application: if you read your data in share mode
and do not see a duplicate for a row you are going to insert,
then you can safely insert your row and know that the next-key
lock set on the successor of your row during the read will prevent
anyone meanwhile inserting a duplicate for your row. Thus the next-key
locking allows you to 'lock' the non-existence of something in your
table.

@node InnoDB Locks set, InnoDB Deadlock detection, InnoDB Next-key locking, InnoDB transaction model
@subsubsection Locks Set by Different SQL Statements in InnoDB

@itemize @bullet
@item
@code{SELECT ... FROM ...} : this is a consistent read, reading a
snapshot of the database and setting no locks.
@item
@code{SELECT ... FROM ... LOCK IN SHARE MODE} : sets shared next-key locks
on all index records the read encounters.
@item
@code{SELECT ... FROM ... FOR UPDATE} : sets exclusive next-key locks
on all index records the read encounters.
@item
@code{INSERT INTO ... VALUES (...)} : sets an exclusive lock
on the inserted row; note that this lock is not a next-key lock
and does not prevent other users from inserting to the gap before the
inserted row. If a duplicate key error occurs, sets a shared lock
on the duplicate index record.
@item
@code{INSERT INTO T SELECT ... FROM S WHERE ...} sets an exclusive
(non-next-key) lock on each row inserted into @code{T}. Does
the search on @code{S} as a consistent read, but sets shared next-key
locks on @code{S} if the MySQL logging is on. InnoDB has to set
locks in the latter case because in roll-forward recovery from a
backup every SQL statement has to be executed in exactly the same
way as it was done originally.
@item
@code{CREATE TABLE ... SELECT ...} performs the @code{SELECT}
as a consistent read or with shared locks, like in the previous
item.
@item
@code{REPLACE} is done like an insert if there is no collision
on a unique key. Otherwise, an exclusive next-key lock is placed
on the row which has to be updated.
@item
@code{UPDATE ... SET ... WHERE ...} : sets an exclusive next-key
lock on every record the search encounters.
@item
@code{DELETE FROM ... WHERE ...} : sets an exclusive next-key
lock on every record the search encounters.
@item
If a @code{FOREIGN KEY} constraint is defined on a table,
any insert, update, or delete which requires checking of the constraint
condition sets shared record level locks on the records it
looks at to check the constraint. Also in the case where the
constraint fails, InnoDB sets these locks.
@item
@code{LOCK TABLES ... } : sets table locks. In the implementation
the MySQL layer of code sets these locks. The automatic deadlock detection
of InnoDB cannot detect deadlocks where such table locks are involved:
see the following section.
Also, since MySQL does know about row level locks,
it is possible that you
get a table lock on a table where another user currently has row level
locks. But that does not put transaction integerity into danger.
@xref{InnoDB restrictions}.
@end itemize


@node InnoDB Deadlock detection, InnoDB Consistent read example, InnoDB Locks set, InnoDB transaction model
@subsubsection Deadlock Detection and Rollback

InnoDB automatically detects a deadlock of transactions and rolls
back the transaction whose lock request was the last one to build
a deadlock, that is, a cycle in the waits-for graph of transactions.
InnoDB cannot detect deadlocks where a lock set by a MySQL
@code{LOCK TABLES} statement is involved, or if a lock set
in another table handler than InnoDB is involved. You have to resolve
these situations using @code{innodb_lock_wait_timeout} set in
@file{my.cnf}.

When InnoDB performs a complete rollback of a transaction, all the
locks of the transaction are released. However, if just a single SQL
statement is rolled back as a result of an error, some of the locks
set by the SQL statement may be preserved. This is because InnoDB
stores row locks in a format where it cannot afterwards know which was
set by which SQL statement.

@node InnoDB Consistent read example,  , InnoDB Deadlock detection, InnoDB transaction model
@subsubsection An Example of How the Consistent Read Works in InnoDB

When you issue a consistent read, that is, an ordinary @code{SELECT}
statement, InnoDB will give your transaction a timepoint according
to which your query sees the database. Thus, if transaction B deletes
a row and commits after your timepoint was assigned, then you will
not see the row deleted. Similarly with inserts and updates.

You can advance your timepoint by committing your transaction
and then doing another @code{SELECT}.

This is called multi-versioned concurrency control.

@example
                  User A                 User B

              SET AUTOCOMMIT=0;      SET AUTOCOMMIT=0;
time
|             SELECT * FROM t;
|             empty set
|                                    INSERT INTO t VALUES (1, 2);
|
v             SELECT * FROM t;
              empty set
                                     COMMIT;

              SELECT * FROM t;
              empty set;

              COMMIT;

              SELECT * FROM t;
              ---------------------
              |    1    |    2    |
              ---------------------
@end example

Thus user A sees the row inserted by B only when B has committed the
insert, and A has committed his own transaction so that the timepoint
is advanced past the commit of B.

If you want to see the ``freshest'' state of the database, you should
use a locking read:

@example
SELECT * FROM t LOCK IN SHARE MODE;
@end example


@subsection Performance Tuning Tips

@strong{1.}
If the Unix @file{top} or the Windows @file{Task Manager} shows that
the CPU usage percentage with your workload is less than 70 %,
your workload is probably
disk-bound. Maybe you are making too many transaction commits, or the
buffer pool is too small.
Making the buffer pool bigger can help, but do not set
it bigger than 80 % of physical memory.

@strong{2.}
Wrap several modifications into one transaction. InnoDB must
flush the log to disk at each transaction commit, if that transaction
made modifications to the database. Since the rotation speed of a disk
is typically
at most 167 revolutions/second, that constrains the number of commits
to the same 167/second if the disk does not fool the operating system.

@strong{3.}
If you can afford the loss of some latest committed transactions, you can
set the @file{my.cnf} parameter @code{innodb_flush_log_at_trx_commit}
to zero. InnoDB tries to flush the log anyway once in a second,
though the flush is not guaranteed.

@strong{4.}
Make your log files big, even as big as the buffer pool. When InnoDB
has written the log files full, it has to write the modified contents
of the buffer pool to disk in a checkpoint. Small log files will cause many
unnecessary disk writes. The drawback in big log files is that recovery
time will be longer.

@strong{5.}
Also the log buffer should be quite big, say 8 MB.

@strong{6.} (Relevant from 3.23.39 up.)
In some versions of Linux and Unix, flushing files to disk with the Unix
@code{fdatasync} and other similar methods is surprisingly slow.
The default method InnoDB uses is the @code{fdatasync} function.
If you are not satisfied with the database write performance, you may
try setting @code{innodb_flush_method} in @file{my.cnf}
to @code{O_DSYNC}, though O_DSYNC seems to be slower on most systems.

@strong{7.} In importing data to InnoDB, make sure that MySQL does not have
@code{autocommit=1} on. Then every insert requires a log flush to disk.
Put before your plain SQL import file line

@example
SET AUTOCOMMIT=0;
@end example

and after it

@example
COMMIT;
@end example

If you use the @file{mysqldump} option @code{--opt}, you will get dump
files which are fast to import also to an InnoDB table, even without wrapping
them to the above @code{SET AUTOCOMMIT=0; ... COMMIT;} wrappers.

@strong{8.}
Beware of big rollbacks of mass inserts: InnoDB uses the insert buffer
to save disk I/O in inserts, but in a corresponding rollback no such
mechanism is used. A disk-bound rollback can take 30 times the time
of the corresponding insert. Killing the database process will not
help because the rollback will start again at the database startup. The
only way to get rid of a runaway rollback is to increase the buffer pool
so that the rollback becomes CPU-bound and runs fast, or delete the whole
InnoDB database.

@strong{9.}
Beware also of other big disk-bound operations.
Use @code{DROP TABLE} or @code{TRUNCATE} (from MySQL-4.0 up) to empty a
table, not @code{DELETE FROM yourtable}.

@strong{10.}
Use the multi-line @code{INSERT} to reduce
communication overhead between the client and the server if you need
to insert many rows:

@example
INSERT INTO yourtable VALUES (1, 2), (5, 5);
@end example

This tip is of course valid for inserts into any table type, not just InnoDB.

@subsubsection The InnoDB Monitor

Starting from version 3.23.41 InnoDB includes the InnoDB
Monitor which prints information on the InnoDB internal state.
When swithed on, InnoDB Monitor
will make the MySQL server @file{mysqld} to print data
(note: the MySQL client will not print anything)
to the standard
output about once every 15 seconds. This data is useful in
performance tuning.
On Windows you must start @code{mysqld-max}
from a MS-DOS prompt
with the @code{--standalone --console}
options to direct the output to the MS-DOS prompt
window.

There is a separate @code{innodb_lock_monitor} which
prints the same information as @code{innodb_monitor}
plus information on locks set by each transaction.

The printed information includes data on:
@itemize @bullet
@item
lock waits of a transactions,
@item
semaphore waits of threads,
@item
pending file I/O requests,
@item
buffer pool statistics, and
@item
purge and insert buffer merge activity of the main thread
of InnoDB.
@end itemize

You can start InnoDB Monitor through the following
SQL command:

@example
CREATE TABLE innodb_monitor(a int) type = innodb;
@end example

and stop it by

@example
DROP TABLE innodb_monitor;
@end example

The @code{CREATE TABLE} syntax is just a way to pass a command
to the InnoDB engine through the MySQL SQL parser: the created
table is not relevant at all for InnoDB Monitor. If you shut down
the database when the monitor is running, and you want to start
the monitor again, you have to drop the
table before you can issue a new @code{CREATE TABLE}
to start the monitor.
This syntax may change in a future release.


A sample output of the InnoDB Monitor:

@example
================================
010809 18:45:06 INNODB MONITOR OUTPUT
================================
--------------------------
LOCKS HELD BY TRANSACTIONS
--------------------------
LOCK INFO:
Number of locks in the record hash table 1294
LOCKS FOR TRANSACTION ID 0 579342744
TABLE LOCK table test/mytable trx id 0 582333343 lock_mode IX

RECORD LOCKS space id 0 page no 12758 n bits 104 table test/mytable index
PRIMARY trx id 0 582333343 lock_mode X
Record lock, heap no 2 PHYSICAL RECORD: n_fields 74; 1-byte offs FALSE;
info bits 0
 0: len 4; hex 0001a801; asc ;; 1: len 6; hex 000022b5b39f; asc ";;
 2: len 7; hex 000002001e03ec; asc ;; 3: len 4; hex 00000001;
...
-----------------------------------------------
CURRENT SEMAPHORES RESERVED AND SEMAPHORE WAITS
-----------------------------------------------
SYNC INFO:
Sorry, cannot give mutex list info in non-debug version!
Sorry, cannot give rw-lock list info in non-debug version!
-----------------------------------------------------
SYNC ARRAY INFO: reservation count 6041054, signal count 2913432
4a239430 waited for by thread 49627477 op. S-LOCK file NOT KNOWN line 0
Mut ex 0 sp 5530989 r 62038708 sys 2155035;
rws 0 8257574 8025336; rwx 0 1121090 1848344
-----------------------------------------------------
CURRENT PENDING FILE I/O'S
--------------------------
Pending normal aio reads:
Reserved slot, messages 40157658 4a4a40b8
Reserved slot, messages 40157658 4a477e28
...
Reserved slot, messages 40157658 4a4424a8
Reserved slot, messages 40157658 4a39ea38
Total of 36 reserved aio slots
Pending aio writes:
Total of 0 reserved aio slots
Pending insert buffer aio reads:
Total of 0 reserved aio slots
Pending log writes or reads:
Reserved slot, messages 40158c98 40157f98
Total of 1 reserved aio slots
Pending synchronous reads or writes:
Total of 0 reserved aio slots
-----------
BUFFER POOL
-----------
LRU list length 8034
Free list length 0
Flush list length 999
Buffer pool size in pages 8192
Pending reads 39
Pending writes: LRU 0, flush list 0, single page 0
Pages read 31383918, created 51310, written 2985115
----------------------------
END OF INNODB MONITOR OUTPUT
============================
010809 18:45:22 InnoDB starts purge
010809 18:45:22 InnoDB purged 0 pages
@end example

Some notes on the output:

@itemize @bullet
@item
If the section LOCKS HELD
BY TRANSACTIONS reports lock waits, then your application
may have lock contention. The output can also help to
trace reasons for transaction deadlocks.
@item
Section SYNC INFO will report reserved semaphores
if you compile InnoDB with @code{UNIV_SYNC_DEBUG}
defined in @file{univ.i}.
@item
Section SYNC ARRAY INFO reports threads waiting
for a semaphore and statistics on how many times
threads have needed a spin or a wait on a mutex or
a rw-lock semaphore. A big number of threads waiting
for semaphores may be a result of disk I/O, or
contention problems inside InnoDB. Contention can be
due to heavy parallelism of queries, or problems in
operating system thread scheduling.
@item
Section CURRENT PENDING FILE I/O'S lists pending
file I/O requests. A large number of these indicates
that the workload is disk I/O-bound.
@item
Section BUFFER POOL gives you statistics
on pages read and written. You can calculate from these
numbers how many datafile I/Os your queries are
currently doing.
@end itemize

@node Implementation, Table and index, InnoDB transaction model, InnoDB
@subsection Implementation of Multi-versioning

Since InnoDB is a multi-versioned database, it must keep information
of old versions of rows in the tablespace. This information is stored
in a data structure we call a rollback segment after an analogous
data structure in Oracle.

InnoDB internally adds two fields to each row stored in the database.
A 6-byte field tells the transaction identifier for the last
transaction which inserted or updated the row. Also a deletion
is internally treated as an update where a special bit in the row
is set to mark it as deleted. Each row also contains a 7-byte
field called the roll pointer. The roll pointer points to an
undo log record written to the rollback segment. If the row was
updated, then the undo log record contains the information necessary
to rebuild the content of the row before it was updated.

InnoDB uses the information in the rollback segment to perform the
undo operations needed in a transaction rollback. It also uses the
information to build earlier versions of a row for a consistent
read.

Undo logs in the rollback segment are divided into insert and update
undo logs. Insert undo logs are only needed in transaction rollback
and can be discarded as soon as the transaction commits. Update undo logs
are used also in consistent reads, and they can be discarded only after
there is no transaction present for which InnoDB has assigned
a snapshot that in a consistent read could need the information
in the update undo log to build an earlier version of a database
row.

You must remember to commit your transactions regularly,
also those transactions which only issue consistent reads.
Otherwise
InnoDB cannot discard data from the update undo logs, and the
rollback segment may grow too big, filling up your tablespace.

The physical size of an undo log record in the rollback segment
is typically smaller than the corresponding inserted or updated
row. You can use this information to calculate the space need
for your rollback segment.

In our multi-versioning scheme a row is not physically removed from
the database immediately when you delete it with an SQL statement.
Only  when InnoDB can discard the update undo log record written for
the deletion, it can also physically remove the corresponding row and
its index records from the database. This removal operation is
called a purge, and it is quite fast, usually taking the same order of
time as the SQL statement which did the deletion.


@node Table and index, File space management, Implementation, InnoDB
@subsection Table and Index Structures

MySQL stores its data dictionary information of tables
in @file{.frm}
files in database directories. But every InnoDB type table
also has its own entry in InnoDB internal data dictionaries
inside the tablespace. When MySQL drops a table or a database,
it has to delete both a @file{.frm} file or files, and
the corresponding entries inside the InnoDB data dictionary.
This is the reason why you cannot move InnoDB tables between
databases simply by moving the @file{.frm} files, and why
@code{DROP DATABASE} did not work for InnoDB type tables
in MySQL versions <= 3.23.43.

Every InnoDB table has a special index called the clustered index
where the data of the rows is stored. If you define a
@code{PRIMARY KEY} on your table, then the index of the primary key
will be the clustered index.

If you do not define a primary key for
your table, InnoDB will internally generate a clustered index
where the rows are ordered by the row id InnoDB assigns
to the rows in such a table. The row id is a 6-byte field which
monotonically increases as new rows are inserted. Thus the rows
ordered by the row id will be physically in the insertion order.

Accessing a row through the clustered index is fast, because
the row data will be on the same page where the index search
leads us. In many databases the data is traditionally stored on a different
page from the index record. If a table is large, the clustered
index architecture often saves a disk I/O when compared to the
traditional solution.

The records in non-clustered indexes (we also call them secondary indexes),
in InnoDB contain the primary key value for the row. InnoDB
uses this primary key value to search for the row from the clustered
index. Note that if the primary key is long, the secondary indexes
will use more space.

@menu
* InnoDB physical structure::   Physical Structure of an Index
* InnoDB Insert buffering::     Insert Buffering
* InnoDB Adaptive hash::        Adaptive Hash Indexes
* InnoDB Physical record::      Physical Record Structure
@end menu


@node InnoDB physical structure, InnoDB Insert buffering, Table and index, Table and index
@subsubsection Physical Structure of an Index

All indexes in InnoDB are B-trees where the index records are
stored in the leaf pages of the tree. The default size of an index
page is 16 kB. When new records are inserted, InnoDB tries to
leave 1 / 16 of the page free for future insertions and updates
of the index records.

If index records are inserted in a sequential (ascending or descending)
order, the resulting index pages will be about 15/16 full.
If records are inserted in a random order, then the pages will be
1/2 - 15/16 full. If the fillfactor of an index page drops below 1/2,
InnoDB will try to contract the index tree to free the page.


@node InnoDB Insert buffering, InnoDB Adaptive hash, InnoDB physical structure, Table and index
@subsubsection Insert Buffering

It is a common situation in a database application that the
primary key is a unique identifier and new rows are inserted in the
ascending order of the primary key. Thus the insertions to the
clustered index do not require random reads from a disk.

On the other hand, secondary indexes are usually non-unique and
insertions happen in a relatively random order into secondary indexes.
This would cause a lot of random disk I/Os without a special mechanism
used in InnoDB.

If an index record should be inserted to a non-unique secondary index,
InnoDB checks if the secondary index page is already in the buffer
pool. If that is the case, InnoDB will do the insertion directly to
the index page. But, if the index page is not found from the buffer
pool, InnoDB inserts the record to a special insert buffer structure.
The insert buffer is kept so small that it entirely fits in the buffer
pool, and insertions can be made to it very fast.

The insert buffer is periodically merged to the secondary index
trees in the database. Often we can merge several insertions on the
same page in of the index tree, and hence save disk I/Os.
It has been measured that the insert buffer can speed up insertions
to a table up to 15 times.


@node InnoDB Adaptive hash, InnoDB Physical record, InnoDB Insert buffering, Table and index
@subsubsection Adaptive Hash Indexes

If a database fits almost entirely in main memory, then the fastest way
to perform queries on it is to use hash indexes. InnoDB has an
automatic mechanism which monitors index searches made to the indexes
defined for a table, and if InnoDB notices that queries could
benefit from building of a hash index, such an index is automatically
built.

But note that the hash index is always built based on an existing
B-tree index on the table. InnoDB can build a hash index on a prefix
of any length of the key defined for the B-tree, depending on
what search pattern InnoDB observes on the B-tree index.
A hash index can be partial: it is not required that the whole
B-tree index is cached in the buffer pool. InnoDB will build
hash indexes on demand to those pages of the index which are
often accessed.

In a sense, through the adaptive hash index mechanism InnoDB adapts itself
to ample main memory, coming closer to the architecture of main memory
databases.


@node InnoDB Physical record,  , InnoDB Adaptive hash, Table and index
@subsubsection Physical Record Structure

@itemize @bullet
@item
Each index record in InnoDB contains a header of 6 bytes. The header
is used to link consecutive records together, and also in the row level
locking.
@item
Records in the clustered index contain fields for all user-defined
columns. In addition, there is a 6-byte field for the transaction id
and a 7-byte field for the roll pointer.
@item
If the user has not defined a primary key for a table, then each clustered
index record contains also a 6-byte row id field.
@item
Each secondary index record contains also all the fields defined
for the clustered index key.
@item
A record contains also a pointer to each field of the record.
If the total length of the fields in a record is < 128 bytes, then
the pointer is 1 byte, else 2 bytes.
@end itemize

@subsubsection How an Auto-increment Column Works in InnoDB

After a database startup, when a user first does an insert to a
table @code{T}
where an auto-increment column has been defined, and the user does not provide
an explicit value for the column, then InnoDB executes @code{SELECT
MAX(auto-inc-column) FROM T}, and assigns that value incremented
by one to the column and the auto-increment counter of the table.
We say that
the auto-increment counter for table @code{T} has been initialised.

InnoDB follows the same procedure in initializing the auto-increment counter
for a freshly created table.

Note that if the user specifies in an insert the value 0 to the auto-increment
column, then InnoDB treats the row like the value would not have been
specified.

After the auto-increment counter has been initialised, if a user inserts
a row where he explicitly specifies the column value, and the value is bigger
than the current counter value, then the counter is set to the specified
column value. If the user does not explicitly specify a value, then InnoDB
increments the counter by one and assigns its new value to the column.

The auto-increment mechanism, when assigning values from the counter,
bypasses locking and transaction handling. Therefore you may also get
gaps in the number sequence if you roll back transactions which have
got numbers from the counter.

The behavior of auto-increment is not defined if a user gives a negative
value to the column or if the value becomes bigger than the maximum
integer that can be stored in the specified integer type.


@node File space management, Error handling, Table and index, InnoDB
@subsection File Space Management and Disk I/O

@menu
* InnoDB Disk I/O::             Disk I/O
* InnoDB File space::           File Space Management
* InnoDB File Defragmenting::   Defragmenting a Table
@end menu


@node InnoDB Disk I/O, InnoDB File space, File space management, File space management
@subsubsection Disk I/O

In disk I/O InnoDB uses asynchronous I/O. On Windows NT
it uses the native asynchronous I/O provided by the operating system.
On Unix, InnoDB uses simulated asynchronous I/O built
into InnoDB: InnoDB creates a number of I/O threads to take care
of I/O operations, such as read-ahead. In a future version we will
add support for simulated aio on Windows NT and native aio on those
versions of Unix which have one.

On Windows NT InnoDB uses non-buffered I/O. That means that the disk
pages InnoDB reads or writes are not buffered in the operating system
file cache. This saves some memory bandwidth.

Starting from 3.23.41 InnoDB uses a novel file flush technique
called doublewrite.
It adds safety to crash recovery after an operating system crash
or a power outage, and improves performance on most Unix flavors
by reducing the need for fsync operations.

Doublewrite means that InnoDB before writing pages to a datafile
first writes them to a contiguous tablespace area called the
doublewrite buffer. Only after the write and the flush to the doublewrite
buffer has completed, InnoDB writes the pages to their proper
positions in the datafile. If the operating system crashes in the
middle of a page write, InnoDB will in recovery find a good
copy of the page from the doublewrite buffer.

Starting from 3.23.41
you can also use a raw disk partition as a datafile, though this has
not been tested yet. When you create a new datafile you have
to put the keyword @code{newraw} immediately after the data
file-size in @code{innodb_data_file_path}. The partition must be
>= than you specify as the size. Note that 1M in InnoDB is
1024 x 1024 bytes, while in disk specifications 1 MB usually means
1000 000 bytes.

@example
innodb_data_file_path=hdd1:5Gnewraw;hdd2:2Gnewraw
@end example

When you start the database again you @strong{must} change the keyword
to @code{raw}. Otherwise, InnoDB will write over your
partition!

@example
innodb_data_file_path=hdd1:5Graw;hdd2:2Graw
@end example

By using a raw disk you can on some Unixes perform unbuffered I/O.

There are two read-ahead heuristics in InnoDB: sequential read-ahead
and random read-ahead. In sequential read-ahead InnoDB notices that
the access pattern to a segment in the tablespace is sequential.
Then InnoDB will post in advance a batch of reads of database pages to the
I/O system. In random read-ahead InnoDB notices that some area
in a tablespace seems to be in the process of being
fully read into the buffer pool. Then InnoDB posts the remaining
reads to the I/O system.


@node InnoDB File space, InnoDB File Defragmenting, InnoDB Disk I/O, File space management
@subsubsection File Space Management

The datafiles you define in the configuration file form the tablespace
of InnoDB. The files are simply catenated to form the tablespace,
there is no striping in use.
Currently you cannot directly instruct where the space is allocated
for your tables, except by using the following fact: from a newly created
tablespace InnoDB will allocate space starting from the low end.

The tablespace consists of database pages whose default size is 16 kB.
The pages are grouped into extents of 64 consecutive pages. The 'files' inside
a tablespace are called segments in InnoDB. The name of the rollback
segment is somewhat misleading because it actually contains many
segments in the tablespace.

For each index in InnoDB we allocate two segments: one is for non-leaf
nodes of the B-tree, the other is for the leaf nodes. The idea here is
to achieve better sequentiality for the leaf nodes, which contain the
data.

When a segment grows inside the tablespace, InnoDB allocates the
first 32 pages to it individually. After that InnoDB starts
to allocate whole extents to the segment.
InnoDB can add to a large segment up to 4 extents at a time to ensure
good sequentiality of data.

Some pages in the tablespace contain bitmaps of other pages, and
therefore a few extents in an InnoDB tablespace cannot be
allocated to segments as a whole, but only as individual pages.

When you issue a query @code{SHOW TABLE STATUS FROM ... LIKE ...}
to ask for available free space in the tablespace, InnoDB will
report the extents which are definitely free in the tablespace.
InnoDB always reserves some extents for clean-up and other internal
purposes; these reserved extents are not included in the free space.

When you delete data from a table, InnoDB will contract the corresponding
B-tree indexes. It depends on the pattern of deletes if that frees
individual pages or extents to the tablespace, so that the freed
space is available for other users. Dropping a table or deleting
all rows from it is guaranteed to release the space to other users,
but remember that deleted rows can be physically removed only in a
purge operation after they are no longer needed in transaction rollback or
consistent read.


@node InnoDB File Defragmenting,  , InnoDB File space, File space management
@subsubsection Defragmenting a Table

If there are random insertions or deletions
in the indexes of a table, the indexes
may become fragmented. By fragmentation we mean that the physical ordering
of the index pages on the disk is not close to the alphabetical ordering
of the records on the pages, or that there are many unused pages in the
64-page blocks which were allocated to the index.

It can speed up index scans if you
periodically use @code{mysqldump} to dump the table to
a text file, drop the table, and reload it from the dump.
Another way to do the defragmenting is to @code{ALTER} the table type to
@code{MyISAM} and back to @code{InnoDB} again.
Note that a @code{MyISAM} table must fit in a single file
on your operating system.

If the insertions to and index are always ascending and
records are deleted only from the end, then the file space management
algorithm of InnoDB guarantees that fragmentation in the index will
not occur.


@node Error handling, InnoDB restrictions, File space management, InnoDB
@subsection Error Handling

The error handling in InnoDB is not always the same as
specified in the ANSI SQL standards. According to the ANSI
standard, any error during an SQL statement should cause the
rollback of that statement. InnoDB sometimes rolls back only
part of the statement, or the whole transaction.
The following list specifies the error handling of InnoDB.

@itemize @bullet
@item
If you run out of file space in the tablespace,
you will get the MySQL @code{'Table is full'} error
and InnoDB rolls back the SQL statement.
@item
A transaction deadlock or a timeout in a lock wait make InnoDB
to roll back
the whole transaction.
@item
A duplicate key error only rolls back the insert of that particular row,
even in a statement like @code{INSERT INTO ... SELECT ...}.
This will probably change so that the SQL statement will be rolled
back if you have not specified the @code{IGNORE} option in your
statement.
@item
A 'row too long' error rolls back the SQL statement.
@item
Other errors are mostly detected by the MySQL layer of code, and
they roll back the corresponding SQL statement.
@end itemize


@node InnoDB restrictions, InnoDB contact information, Error handling, InnoDB
@subsection Restrictions on InnoDB Tables

@itemize @bullet

@item
@strong{Warning}: do @strong{NOT} convert MySQL system tables from
MyISAM TO InnoDB tables!  This is not supported; if you do this MySQL
will not restart until you restore the old system tables from a backup
or re-generate them with the mysql_install_db script.

@item
@code{SHOW TABLE STATUS} does not give accurate statistics
on InnoDB tables, except for the physical size reserved by the table.
The row count is only a rough estimate used in SQL optimisation.

@item
If you try to create an unique index on a prefix of a column you will get an
error:

@example
CREATE TABLE T (A CHAR(20), B INT, UNIQUE (A(5))) TYPE = InnoDB;
@end example

If you create a non-unique index on a prefix of a column, InnoDB will
create an index over the whole column.
@item
@code{INSERT DELAYED} is not supported for InnoDB tables.
@item
The MySQL @code{LOCK TABLES} operation does not know of InnoDB
row level locks set in already completed SQL statements: this means that
you can get a table lock on a table even if there still exist transactions
of other users which have row level locks on the same table. Thus
your operations on the table may have to wait if they collide with
these locks of other users. Also a deadlock is possible. However,
this does not endanger transaction integrity, because the row level
locks set by InnoDB will always take care of the integrity.
Also, a table lock prevents other transactions from acquiring more
row level locks (in a conflicting lock mode) on the table.
@item
You cannot have a key on a @code{BLOB} or @code{TEXT} column.
@item
A table cannot contain more than 1000 columns.
@item
@code{DELETE FROM TABLE} does not regenerate the table but instead
deletes all rows, one by one, which is not that fast. In future versions
of MySQL you can use @code{TRUNCATE} which is fast.
@item
The default database page size in InnoDB is 16 kB. By recompiling the
code one can set it from 8 kB to 64 kB.
The maximun row length is slightly less than half of a database page
in versions <= 3.23.40 of InnoDB. Starting from source
release 3.23.41 BLOB and
TEXT columns are allowed to be < 4 GB, the total row length must also be
< 4 GB. InnoDB does not store fields whose size is <= 128 bytes on separate
pages. After InnoDB has modified the row by storing long fields on
separate pages, the remaining length of the row must be less
than half a database page. The maximun key length is 7000 bytes.
@item
On some operating systems datafiles must be < 2 GB. The combined
size of log files must be < 4 GB on 32-bit computers.
@item
The maximum tablespace size is 4 billion database pages. This is also
the maximum size for a table. The minimum tablespace size is 10 MB.
@end itemize


@node InnoDB contact information,  , InnoDB restrictions, InnoDB
@subsection InnoDB Contact Information

Contact information of Innobase Oy, producer of the InnoDB engine.
Web site: @uref{http://www.innodb.com/}.
E-mail: @email{Heikki.Tuuri@@innodb.com}

@example
phone: 358-9-6969 3250 (office) 358-40-5617367 (mobile)
Innobase Oy Inc.
World Trade Center Helsinki
Aleksanterinkatu 17
P.O.Box 800
00101 Helsinki
Finland
@end example


@node BDB,  , InnoDB, Table types
@section BDB or Berkeley_DB Tables

@cindex tables, @code{BDB}
@cindex tables, @code{Berkeley DB}

@menu
* BDB overview::                Overview of BDB Tables
* BDB install::                 Installing BDB
* BDB start::                   BDB startup options
* BDB characteristics::         Characteristics of @code{BDB} tables:
* BDB TODO::                    Things we need to fix for BDB in the near future:
* BDB portability::             Operating systems supported by @code{BDB}
* BDB restrictions::            Restrictions on BDB Tables
* BDB errors::                  Errors That May Occur When Using BDB Tables
@end menu

@node BDB overview, BDB install, BDB, BDB
@subsection Overview of BDB Tables

Support for BDB tables is included in the MySQL source distribution
starting from Version 3.23.34 and is activated in the MySQL-Max
binary.

BerkeleyDB, available at @uref{http://www.sleepycat.com/} has provided
MySQL with a transactional table handler.  By using BerkeleyDB
tables, your tables may have a greater chance of surviving crashes, and also
provides @code{COMMIT} and @code{ROLLBACK} on transactions.  The
MySQL source distribution comes with a BDB distribution that has a
couple of small patches to make it work more smoothly with MySQL.
You can't use a non-patched @code{BDB} version with MySQL.

We at MySQL AB are working in close cooperation with Sleepycat to
keep the quality of the MySQL/BDB interface high.

When it comes to supporting BDB tables, we are committed to help our
users to locate the problem and help creating a reproducable test case
for any problems involving BDB tables.  Any such test case will be
forwarded to Sleepycat who in turn will help us find and fix the
problem.  As this is a two stage operation, any problems with BDB tables
may take a little longer for us to fix than for other table handlers.
However, as the BerkeleyDB code itself has been used by many other
applications than MySQL, we don't envision any big problems with
this.  @xref{Support}.


@node BDB install, BDB start, BDB overview, BDB
@subsection Installing BDB

If you have downloaded a binary version of MySQL that includes
support for BerkeleyDB, simply follow the instructions for installing a
binary version of MySQL.
@xref{Installing binary}.  @xref{mysqld-max, , @code{mysqld-max}}.

To compile MySQL with Berkeley DB support, download MySQL
Version 3.23.34 or newer and configure @code{MySQL} with the
@code{--with-berkeley-db} option.  @xref{Installing source}.

@example
cd /path/to/source/of/mysql-3.23.34
./configure --with-berkeley-db
@end example

Please refer to the manual provided with the @code{BDB} distribution for
more updated information.

Even though Berkeley DB is in itself very tested and reliable,
the MySQL interface is still considered beta quality.
We are actively improving and optimising it to get it stable very
soon.


@node BDB start, BDB characteristics, BDB install, BDB
@subsection BDB startup options

If you are running with @code{AUTOCOMMIT=0} then your changes in @code{BDB}
tables will not be updated until you execute @code{COMMIT}.  Instead of commit
you can execute @code{ROLLBACK} to forget your changes. @xref{COMMIT}.

If you are running with @code{AUTOCOMMIT=1} (the default), your changes
will be committed immediately.  You can start an extended transaction with
the @code{BEGIN WORK} SQL command, after which your changes will not be
committed until you execute @code{COMMIT} (or decide to @code{ROLLBACK}
the changes).

The following options to @code{mysqld} can be used to change the behavior of
BDB tables:

@multitable @columnfractions .25 .55
@item @strong{Option} @tab @strong{Description}
@item @code{--bdb-home=directory} @tab  Base directory for BDB tables. This should be the same directory you use for --datadir.
@item @code{--bdb-lock-detect=#} @tab  Berkeley lock detect. One of (DEFAULT, OLDEST, RANDOM, or YOUNGEST).
@item @code{--bdb-logdir=directory} @tab Berkeley DB log file directory.
@item @code{--bdb-no-sync} @tab Don't synchronously flush logs.
@item @code{--bdb-no-recover} @tab Don't start Berkeley DB in recover mode.
@item @code{--bdb-shared-data} @tab Start Berkeley DB in multi-process mode (Don't use @code{DB_PRIVATE} when initialising Berkeley DB)
@item @code{--bdb-tmpdir=directory} @tab Berkeley DB tempfile name.
@item @code{--skip-bdb} @tab Don't use berkeley db.
@item @code{-O bdb_max_lock=1000} @tab Set the maximum number of locks possible. @xref{SHOW VARIABLES}.
@end multitable

If you use @code{--skip-bdb}, MySQL will not initialise the
Berkeley DB library and this will save a lot of memory. Of course,
you cannot use @code{BDB} tables if you are using this option.

Normally you should start @code{mysqld} without @code{--bdb-no-recover} if you
intend to use BDB tables.  This may, however, give you problems when you
try to start @code{mysqld} if the BDB log files are corrupted. @xref{Starting
server}.

With @code{bdb_max_lock} you can specify the maximum number of locks
(10000 by default) you can have active on a BDB table. You should
increase this if you get errors of type @code{bdb: Lock table is out of
available locks} or @code{Got error 12 from ...}  when you have do long
transactions or when @code{mysqld} has to examine a lot of rows to
calculate the query.

You may also want to change @code{binlog_cache_size} and
@code{max_binlog_cache_size} if you are using big multi-line transactions.
@xref{COMMIT, , @code{COMMIT}}.


@node BDB characteristics, BDB TODO, BDB start, BDB
@subsection Characteristics of @code{BDB} tables:

@itemize @bullet
@item
To be able to rollback transactions BDB maintain log files.  For maximum
performance you should place these on another disk than your databases
by using the @code{--bdb_log_dir} options.
@item
MySQL performs a checkpoint each time a new BDB log
file is started, and removes any log files that are not needed for
current transactions.  One can also run @code{FLUSH LOGS} at any time
to checkpoint the Berkeley DB tables.

For disaster recovery, one should use table backups plus
MySQL's binary log. @xref{Backup}.

@strong{Warning}: If you delete old log files that are in use, BDB will
not be able to do recovery at all and you may lose data if something
goes wrong.
@item
MySQL requires a @code{PRIMARY KEY} in each BDB table to be
able to refer to previously read rows. If you don't create one,
MySQL will create an maintain a hidden @code{PRIMARY KEY} for
you.  The hidden key has a length of 5 bytes and is incremented for each
insert attempt.
@item
If all columns you access in a @code{BDB} table are part of the same index or
part of the primary key, then MySQL can execute the query
without having to access the actual row.  In a @code{MyISAM} table the
above holds only if the columns are part of the same index.
@item
The @code{PRIMARY KEY} will be faster than any other key, as the
@code{PRIMARY KEY} is stored together with the row data.  As the other keys are
stored as the key data + the @code{PRIMARY KEY}, it's important to keep the
@code{PRIMARY KEY} as short as possible to save disk and get better speed.
@item
@code{LOCK TABLES} works on @code{BDB} tables as with other tables.  If
you don't use @code{LOCK TABLE}, MySQL will issue an internal
multiple-write lock on the table to ensure that the table will be
properly locked if another thread issues a table lock.
@item
Internal locking in @code{BDB} tables is done on page level.
@item
@code{SELECT COUNT(*) FROM table_name} is slow as @code{BDB} tables doesn't
maintain a count of the number of rows in the table.
@item
Scanning is slower than with @code{MyISAM} tables as one has data in BDB
tables stored in B-trees and not in a separate datafile.
@item
The application must always be prepared to handle cases where
any change of a @code{BDB} table may make an automatic rollback and any
read may fail with a deadlock error.
@item
Keys are not compressed to previous keys as with ISAM or MyISAM
tables. In other words, the key information will take a little more
space in @code{BDB} tables compared to MyISAM tables which don't use
@code{PACK_KEYS=0}.
@item
There is often holes in the BDB table to allow you to insert new rows in
the middle of the key tree.  This makes BDB tables somewhat larger than
MyISAM tables.
@item
The optimiser needs to know an approximation of the number of rows in
the table.  MySQL solves this by counting inserts and
maintaining this in a separate segment in each BDB table.  If you don't
issue a lot of @code{DELETE} or @code{ROLLBACK} statements, this number
should be accurate enough for the MySQL optimiser, but as MySQL
only stores the number on close, it may be incorrect if MySQL dies
unexpectedly. It should not be fatal even if this number is not 100 %
correct.  One can update the number of rows by executing @code{ANALYZE
TABLE} or @code{OPTIMIZE TABLE}. @xref{ANALYZE TABLE} . @xref{OPTIMIZE
TABLE}.
@item
If you get full disk with a @code{BDB} table, you will get an error
(probably error 28) and the transaction should roll back.  This is in
contrast with @code{MyISAM} and @code{ISAM} tables where @code{mysqld} will
wait for enough free disk before continuing.
@end itemize


@node BDB TODO, BDB portability, BDB characteristics, BDB
@subsection Things we need to fix for BDB in the near future:

@itemize @bullet
@item
It's very slow to open many BDB tables at the same time. If you are
going to use BDB tables, you should not have a very big table cache
(like >256) and you should use @code{--no-auto-rehash} with the @code{mysql}
client.  We plan to partly fix this in 4.0.
@item
@code{SHOW TABLE STATUS} doesn't yet provide that much information for BDB
tables.
@item
Optimise performance.
@item
Change to not use page locks at all when we are scanning tables.
@end itemize


@node BDB portability, BDB restrictions, BDB TODO, BDB
@subsection Operating systems supported by @code{BDB}

If you after having built MySQL with support for BDB tables get
the following error in the log file when you start @code{mysqld}:

@example
bdb: architecture lacks fast mutexes: applications cannot be threaded
Can't init dtabases
@end example

This means that @code{BDB} tables are not supported for your architecture.
In this case you have to rebuild MySQL without BDB table support.

Note: The following list is not complete; we will update it as we
receive more information about this.

Currently we know that BDB tables works with the following operating
system.

@itemize @bullet
@item
Linux 2.x intel
@item
Solaris SPARC
@item
Caldera (SCO) OpenServer
@item
Caldera (SCO) UnixWare 7.0.1
@end itemize

It doesn't work with the following operating systems:

@itemize @bullet
@item
Linux 2.x Alpha
@item
Max OS X
@end itemize

@node BDB restrictions, BDB errors, BDB portability, BDB
@subsection Restrictions on BDB Tables

Here follows the restrictions you have when using BDB tables:

@itemize @bullet
@item
BDB tables store in the .db file the path to the file as it was created
This was done to be able to detect locks in a multi-user
environment that supports symlinks).

The effect of this is that BDB tables are not movable between directories!
@item
When taking backups of BDB tables, you have to either use
@code{mysqldump} or take a backup of all @code{table_name.db} files and
the BDB log files.  The BDB log files are the files in the base data
directory named @code{log.XXXXXX} (6 digits);
The BDB table handler stores unfinished transactions in the log files
and requires these to be present when @code{mysqld} starts.
@end itemize

@node BDB errors,  , BDB restrictions, BDB
@subsection Errors That May Occur When Using BDB Tables

@itemize @bullet
@item
If you get the following error in the @code{hostname.err log} when
starting @code{mysqld}:

@example
bdb:  Ignoring log file: .../log.XXXXXXXXXX: unsupported log version #
@end example
it means that the new @code{BDB} version doesn't support the old log
file format.  In this case you have to delete all @code{BDB} log BDB
from your database directory (the files that has the format
@code{log.XXXXXXXXXX} ) and restart @code{mysqld}.  We would also
recommend you to do a @code{mysqldump --opt} of your old @code{BDB}
tables, delete the old table and restore the dump.
@item
If you are running in not @code{auto_commit} mode and delete a table you
are using by another thread you may get the following error messages in
the MySQL error file:

@example
001119 23:43:56  bdb:  Missing log fileid entry
001119 23:43:56  bdb:  txn_abort: Log undo failed for LSN:
                       1 3644744: Invalid
@end example

This is not fatal but we don't recommend that you delete tables if you are
not in @code{auto_commit} mode, until this problem is fixed (the fix is
not trivial).
@end itemize




@node Clients, Extending MySQL, Table types, Top
@chapter MySQL APIs

@cindex client tools
@cindex APIs
@cindex @code{mysqlclient} library
@cindex buffer sizes, client
@cindex library, @code{mysqlclient}

@menu
* PHP::                         MySQL PHP API
* Perl::                        MySQL Perl API
* ODBC::                        MySQL ODBC Support
* C::                           MySQL C API
* Cplusplus::                   MySQL C++ APIs
* Java::                        MySQL Java connectivity (JDBC)
* Python::                      MySQL Python APIs
* Tcl::                         MySQL Tcl APIs
* Eiffel::                      MySQL Eiffel wrapper
@end menu

This chapter describes the APIs available for MySQL, where to get
them, and how to use them.  The C API is the most extensively covered, as it
was developed by the MySQL team, and is the basis for most of the
other APIs.


@node PHP, Perl, Clients, Clients
@section MySQL PHP API

@cindex PHP API

PHP is a server-side, HTML-embedded scripting language that may be used to
create dynamic web pages.  It contains support for accessing several
databases, including MySQL.  PHP may be run as a separate program
or compiled as a module for use with the Apache web server.

The distribution and documentation are available at the PHP web site
(@uref{http://www.php.net/}).

@menu
* PHP problems::                Common problems with MySQL and PHP
@end menu

@node PHP problems,  , PHP, PHP
@subsection Common Problems with MySQL and PHP

@itemize @bullet
@item Error: "Maximum Execution Time Exceeded"
This is a PHP limit; go into the @file{php3.ini} file and set the maximum
execution time up from 30 seconds to something higher, as needed.
It is also not a bad idea to double the ram allowed per script to 16MB instead of
8 MB.
@item Error: "Fatal error: Call to unsupported or undefined function mysql_connect() in .."
This means that your PHP version isn't compiled with MySQL support.
You can either compile a dynamic MySQL module and load it into PHP or
recompile PHP with built-in MySQL support.  This is described in
detail in the PHP manual.
@item Error: "undefined reference to `uncompress'"
This means that the client library is compiled with support for a compressed
client/server protocol. The fix is to add @code{-lz} last when linking
with @code{-lmysqlclient}.
@end itemize


@node Perl, ODBC, PHP, Clients
@section MySQL Perl API

@cindex APIs, Perl
@cindex Perl API

This section documents the Perl @code{DBI} interface.  The former interface
was called @code{mysqlperl}.  @code{DBI}/@code{DBD} now is the
recommended Perl interface, so @code{mysqlperl} is obsolete and is not
documented here.

@menu
* DBI with DBD::                @code{DBI} with @code{DBD::mysql}
* Perl DBI Class::              The @code{DBI} interface
* DBI-info::                    More @code{DBI}/@code{DBD} information
@end menu


@node DBI with DBD, Perl DBI Class, Perl, Perl
@subsection @code{DBI} with @code{DBD::mysql}

@cindex @code{DBI} interface

@code{DBI} is a generic interface for many databases. That means that
you can write a script that works with many different database engines
without change. You need a DataBase Driver (DBD) defined for each
database type. For MySQL, this driver is called
@code{DBD::mysql}.

For more information on the Perl5 DBI, please visit the @code{DBI} web
page and read the documentation:
@example
@uref{http://dbi.perl.org/}
@end example
For more information on Object Oriented Programming
(OOP) as defined in Perl5, see the Perl OOP page:
@example
@uref{http://language.perl.com/info/documentation.html}
@end example

Note that if you want to use transactions with Perl, you need to have
@code{Msql-Mysql-modules} version 1.2216 or newer.

Installation instructions for MySQL Perl support are given in
@ref{Perl support}.


@node Perl DBI Class, DBI-info, DBI with DBD, Perl
@subsection The @code{DBI} Interface

@cindex @code{DBI} Perl module

@noindent
@strong{Portable DBI Methods}

@multitable @columnfractions .25 .60
@item @strong{Method} @tab @strong{Description}
@item @code{connect} @tab Establishes a connection to a database server.
@item @code{disconnect} @tab Disconnects from the database server.
@item @code{prepare} @tab Prepares a SQL statement for execution.
@item @code{execute} @tab Executes prepared statements.
@item @code{do}      @tab Prepares and executes a SQL statement.
@item @code{quote}   @tab Quotes string or @code{BLOB} values to be inserted.
@item @code{fetchrow_array} @tab Fetches the next row as an array of fields.
@item @code{fetchrow_arrayref} @tab Fetches next row as a reference array of fields.
@item @code{fetchrow_hashref} @tab Fetches next row as a reference to a hashtable.
@item @code{fetchall_arrayref} @tab Fetches all data as an array of arrays.
@item @code{finish}  @tab Finishes a statement and lets the system free resources.
@item @code{rows}    @tab Returns the number of rows affected.
@item @code{data_sources} @tab Returns an array of databases available on localhost.
@item @code{ChopBlanks} @tab Controls whether @code{fetchrow_*} methods trim spaces.
@item @code{NUM_OF_PARAMS} @tab The number of placeholders in the prepared statement.
@item @code{NULLABLE} @tab Which columns can be @code{NULL}.
@item @code{trace}    @tab Perform tracing for debugging.
@end multitable

@noindent
@strong{MySQL-specific Methods}

@multitable @columnfractions .25 .60
@item @strong{Method} @tab @strong{Description}
@item @code{insertid} @tab The latest @code{AUTO_INCREMENT} value.
@item @code{is_blob} @tab Which columns are @code{BLOB} values.
@item @code{is_key} @tab Which columns are keys.
@item @code{is_num} @tab Which columns are numeric.
@item @code{is_pri_key} @tab Which columns are primary keys.
@item @code{is_not_null} @tab Which columns CANNOT be @code{NULL}. See @code{NULLABLE}.
@item @code{length} @tab Maximum possible column sizes.
@item @code{max_length} @tab Maximum column sizes actually present in result.
@item @code{NAME} @tab Column names.
@item @code{NUM_OF_FIELDS} @tab Number of fields returned.
@item @code{table} @tab Table names in returned set.
@item @code{type} @tab All column types.
@end multitable

The Perl methods are described in more detail in the following sections.
Variables used for method return values have these meanings:

@table @code
@item $dbh
Database handle

@item $sth
Statement handle

@item $rc
Return code (often a status)

@item $rv
Return value (often a row count)
@end table

@noindent
@strong{Portable DBI Methods}

@table @code

@findex DBI->connect()
@findex connect() DBI method
@item connect($data_source, $username, $password)
Use the @code{connect} method to make a database connection to the data
source. The @code{$data_source} value should begin with
@code{DBI:driver_name:}.
Example uses of @code{connect} with the @code{DBD::mysql} driver:
@example
$dbh = DBI->connect("DBI:mysql:$database", $user, $password);
$dbh = DBI->connect("DBI:mysql:$database:$hostname",
                    $user, $password);
$dbh = DBI->connect("DBI:mysql:$database:$hostname:$port",
                    $user, $password);
@end example
If the user name and/or password are undefined, @code{DBI} uses the
values of the @code{DBI_USER} and @code{DBI_PASS} environment variables,
respectively. If you don't specify a hostname, it defaults to
@code{'localhost'}. If you don't specify a port number, it defaults to the
default MySQL port (@value{default_port}).

As of @code{Msql-Mysql-modules} Version 1.2009,
the @code{$data_source} value allows certain modifiers:

@table @code
@item mysql_read_default_file=file_name
Read @file{filename} as an option file.  For information on option files,
see @ref{Option files}.

@item mysql_read_default_group=group_name
The default group when reading an option file is normally the
@code{[client]} group.  By specifying the @code{mysql_read_default_group}
option, the default group becomes the @code{[group_name]} group.

@item mysql_compression=1
Use compressed communication between the client and server (MySQL
Version 3.22.3 or later).

@item mysql_socket=/path/to/socket
Specify the pathname of the Unix socket that is used to connect
to the server (MySQL Version 3.21.15 or later).
@end table

Multiple modifiers may be given; each must be preceded by a semicolon.

For example, if you want to avoid hardcoding the user name and password into
a @code{DBI} script, you can take them from the user's @file{~/.my.cnf}
option file instead by writing your @code{connect} call like this:

@example
$dbh = DBI->connect("DBI:mysql:$database"
                . ";mysql_read_default_file=$ENV@{HOME@}/.my.cnf",
                $user, $password);
@end example

This call will read options defined for the @code{[client]} group in the
option file.  If you wanted to do the same thing but use options specified
for the @code{[perl]} group as well, you could use this:

@example
$dbh = DBI->connect("DBI:mysql:$database"
                . ";mysql_read_default_file=$ENV@{HOME@}/.my.cnf"
                . ";mysql_read_default_group=perl",
                $user, $password);
@end example

@findex DBI->disconnect
@findex disconnect DBI method
@item disconnect
The @code{disconnect} method disconnects the database handle from the database.
This is typically called right before you exit from the program.
Example:
@example
$rc = $dbh->disconnect;
@end example

@findex DBI->prepare()
@findex prepare() DBI method
@item prepare($statement)
Prepares a SQL statement for execution by the database engine
and returns a statement handle @code{($sth)}, which you can use to invoke
the @code{execute} method.
Typically you handle @code{SELECT} statements (and @code{SELECT}-like statements
such as @code{SHOW}, @code{DESCRIBE}, and @code{EXPLAIN}) by means of
@code{prepare} and @code{execute}.
Example:
@example
$sth = $dbh->prepare($statement)
    or die "Can't prepare $statement: $dbh->errstr\n";
@end example

@findex DBI->execute
@findex execute DBI method
@item execute
The @code{execute} method executes a prepared statement. For
non-@code{SELECT} statements, @code{execute} returns the number of rows
affected.  If no rows are affected, @code{execute} returns @code{"0E0"},
which Perl treats as zero but regards as true. If an error occurs,
@code{execute} returns @code{undef}. For @code{SELECT} statements,
@code{execute} only starts the SQL query in the database; you need to use one
of the @code{fetch_*} methods described here to retrieve the data.
Example:
@example
$rv = $sth->execute
          or die "can't execute the query: $sth->errstr;
@end example

@findex DBI->do()
@findex do() DBI method
@item do($statement)
The @code{do} method prepares and executes a SQL statement and returns the
number of rows affected.  If no rows are affected, @code{do} returns
@code{"0E0"}, which Perl treats as zero but regards as true.  This method is
generally used for non-@code{SELECT} statements that cannot be prepared in
advance (due to driver limitations) or that do not need to be executed more
than once (inserts, deletes, etc.). Example:
@example
$rv = $dbh->do($statement)
        or die "Can't execute $statement: $dbh- >errstr\n";
@end example

Generally the 'do' statement is much faster (and is preferable)
than prepare/execute for statements that don't contain parameters.

@findex DBI->quote()
@findex quote() DBI method
@cindex quoting strings
@cindex strings, quoting
@item quote($string)
The @code{quote} method is used to "escape" any special characters contained in
the string and to add the required outer quotation marks.
Example:
@example
$sql = $dbh->quote($string)
@end example

@findex DBI->fetchrow_array
@findex fetchrow_array DBI method
@item fetchrow_array
This method fetches the next row of data and returns it as an array of
field values.  Example:
@example
while(@@row = $sth->fetchrow_array) @{
        print qw($row[0]\t$row[1]\t$row[2]\n);
@}
@end example

@findex DBI->fetchrow_arrayref
@findex fetchrow_arrayref DBI method
@item fetchrow_arrayref
This method fetches the next row of data and returns it as a reference
to an array of field values. Example:
@example
while($row_ref = $sth->fetchrow_arrayref) @{
        print qw($row_ref->[0]\t$row_ref->[1]\t$row_ref->[2]\n);
@}
@end example

@findex DBI->fetchrow_hashref
@findex fetchrow_hashref DBI method
@item fetchrow_hashref
This method fetches a row of data and returns a reference to a hash
table containing field name/value pairs. This method is not nearly as
efficient as using array references as demonstrated above. Example:
@example
while($hash_ref = $sth->fetchrow_hashref) @{
        print qw($hash_ref->@{firstname@}\t$hash_ref->@{lastname@}\t\
                $hash_ref- > title@}\n);
@}
@end example

@findex DBI->fetchall_arrayref
@findex fetchall_arrayref DBI method
@item fetchall_arrayref
This method is used to get all the data (rows) to be returned from the
SQL statement. It returns a reference to an array of references to arrays
for each row. You access or print the data by using a nested
loop. Example:
@example
my $table = $sth->fetchall_arrayref
                or die "$sth->errstr\n";
my($i, $j);
for $i ( 0 .. $#@{$table@} ) @{
        for $j ( 0 .. $#@{$table->[$i]@} ) @{
                print "$table->[$i][$j]\t";
        @}
        print "\n";
@}
@end example

@findex DBI->finish
@findex finish DBI method
@item finish
Indicates that no more data will be fetched from this statement
handle. You call this method to free up the statement handle and any
system resources associated with it.  Example:
@example
$rc = $sth->finish;
@end example

@findex DBI->rows
@findex rows DBI method
@item rows
Returns the number of rows changed (updated, deleted, etc.) by the last
command.  This is usually used after a non-@code{SELECT} @code{execute}
statement.  Example:
@example
$rv = $sth->rows;
@end example

@findex DBI->@{NULLABLE@}
@findex NULLABLE DBI method
@item NULLABLE
Returns a reference to an array of boolean values; for each element of
the array, a value of TRUE indicates that this
column may contain @code{NULL} values.
Example:
@example
$null_possible = $sth->@{NULLABLE@};
@end example

@findex DBI->@{NUM_OF_FIELDS@}
@findex NUM_OF_FIELDS DBI method
@item NUM_OF_FIELDS
This attribute indicates
the number of fields returned by a @code{SELECT} or @code{SHOW FIELDS}
statement.  You may use this for checking whether a statement returned a
result: A zero value indicates a non-@code{SELECT} statement like
@code{INSERT}, @code{DELETE}, or @code{UPDATE}.
Example:
@example
$nr_of_fields = $sth->@{NUM_OF_FIELDS@};
@end example

@findex DBI->data_sources()
@findex data_sources() DBI method
@item data_sources($driver_name)
This method returns an array containing names of databases available to the
MySQL server on the host @code{'localhost'}.
Example:
@example
@@dbs = DBI->data_sources("mysql");
@end example

@findex DBI->@{ChopBlanks@}
@findex ChopBlanks DBI method
@item ChopBlanks
This attribute determines whether the @code{fetchrow_*} methods will chop
leading and trailing blanks from the returned values.
Example:
@example
$sth->@{'ChopBlanks'@} =1;
@end example

@findex DBI->trace
@findex trace DBI method
@item trace($trace_level)
@itemx trace($trace_level, $trace_filename)
The @code{trace} method enables or disables tracing.  When invoked as a
@code{DBI} class method, it affects tracing for all handles.  When invoked as
a database or statement handle method, it affects tracing for the given
handle (and any future children of the handle).  Setting @code{$trace_level}
to 2 provides detailed trace information.  Setting @code{$trace_level} to 0
disables tracing.  Trace output goes to the standard error output by
default.  If @code{$trace_filename} is specified, the file is opened in
append mode and output for @strong{all} traced handles is written to that
file.  Example:
@example
DBI->trace(2);                # trace everything
DBI->trace(2,"/tmp/dbi.out"); # trace everything to
                              # /tmp/dbi.out
$dth->trace(2);               # trace this database handle
$sth->trace(2);               # trace this statement handle
@end example

@tindex @code{DBI_TRACE} environment variable
@tindex environment variable, @code{DBI_TRACE}
You can also enable @code{DBI} tracing by setting the @code{DBI_TRACE}
environment variable.  Setting it to a numeric value is equivalent to calling
@code{DBI->(value)}.  Setting it to a pathname is equivalent to calling
@code{DBI->(2,value)}.

@end table

@noindent
@strong{MySQL-specific Methods}

The methods shown here are MySQL-specific and not part of the
@code{DBI} standard.  Several of them are now deprecated:
@code{is_blob}, @code{is_key}, @code{is_num}, @code{is_pri_key},
@code{is_not_null}, @code{length}, @code{max_length}, and @code{table}.
Where @code{DBI}-standard alternatives exist, they are noted here:

@table @code
@findex DBI->@{insertid@}
@findex insertid DBI method
@tindex AUTO_INCREMENT, using with DBI
@item insertid
If you use the @code{AUTO_INCREMENT} feature of MySQL, the new
auto-incremented values will be stored here.
Example:
@example
$new_id = $sth->@{insertid@};
@end example

As an alternative, you can use @code{$dbh->@{'mysql_insertid'@}}.

@findex DBI->@{is_blob@}
@findex is_blob DBI method
@item is_blob
Returns a reference to an array of boolean values; for each element of the
array, a value of TRUE indicates that the
respective column is a @code{BLOB}.
Example:
@example
$keys = $sth->@{is_blob@};
@end example

@findex DBI->@{is_key@}
@findex is_key DBI method
@item is_key
Returns a reference to an array of boolean values; for each element of the
array, a value of TRUE indicates that the
respective column is a key.
Example:
@example
$keys = $sth->@{is_key@};
@end example

@findex DBI->@{is_num@}
@findex is_num DBI method
@item is_num
Returns a reference to an array of boolean values; for each element of the
array, a value of TRUE indicates that the
respective column contains numeric values.
Example:
@example
$nums = $sth->@{is_num@};
@end example

@findex DBI->@{is_pri_key@}
@findex is_pri_key DBI method
@item is_pri_key
Returns a reference to an array of boolean values; for each element of the
array, a value of TRUE indicates that the respective column is a primary key.
Example:
@example
$pri_keys = $sth->@{is_pri_key@};
@end example

@findex DBI->@{is_not_null@}
@findex is_not_null DBI method
@item is_not_null
Returns a reference to an array of boolean values; for each element of the
array, a value of FALSE indicates that this column may contain @code{NULL}
values.
Example:
@example
$not_nulls = $sth->@{is_not_null@};
@end example

@code{is_not_null} is deprecated; it is preferable to use the
@code{NULLABLE} attribute (described above), because that is a DBI standard.

@findex DBI->@{length@}
@findex length DBI method
@findex DBI->@{max_length@}
@findex max_length DBI method
@item length
@itemx max_length
Each of these methods returns a reference to an array of column sizes.  The
@code{length} array indicates the maximum possible sizes that each column may
be (as declared in the table description).  The @code{max_length} array
indicates the maximum sizes actually present in the result table.  Example:

@example
$lengths = $sth->@{length@};
$max_lengths = $sth->@{max_length@};
@end example

@findex DBI->@{NAME@}
@findex NAME DBI method
@item NAME
Returns a reference to an array of column names.
Example:
@example
$names = $sth->@{NAME@};
@end example

@findex DBI->@{table@}
@findex table DBI method
@item table
Returns a reference to an array of table names.
Example:
@example
$tables = $sth->@{table@};
@end example

@findex DBI->@{type@}
@findex type DBI method
@item type
Returns a reference to an array of column types.
Example:
@example
$types = $sth->@{type@};
@end example

@end table


@node DBI-info,  , Perl DBI Class, Perl
@subsection More @code{DBI}/@code{DBD} Information

@cindex @code{DBI/DBD}

You can use the @code{perldoc} command to get more information about
@code{DBI}.

@example
perldoc DBI
perldoc DBI::FAQ
perldoc DBD::mysql
@end example

You can also use the @code{pod2man}, @code{pod2html}, etc., tools to
translate to other formats.

You can find the latest @code{DBI} information at
the @code{DBI} web page:
@example
@uref{http://www.symbolstone.org/technology/perl/DBI/}
@end example


@node ODBC, C, Perl, Clients
@section MySQL ODBC Support

@cindex ODBC
@cindex Windows
@cindex MyODBC

@menu
* Installing MyODBC::           How to install MyODBC
* ODBC administrator::          How to fill in the various fields in the ODBC administrator program
* MyODBC connect parameters::   Connect parameters for MyODBC
* ODBC Problems::               How to report problems with MySQL ODBC
* MyODBC clients::              Programs known to work with @code{MyODBC}
* ODBC and last_insert_id::     How to get the value of an @code{AUTO_INCREMENT} column in ODBC
* MyODBC bug report::           Reporting problems with MyODBC
@end menu


MySQL provides support for ODBC by means of the @code{MyODBC}
program.  This chapter will teach you how to install @code{MyODBC},
and how to use it.  Here, you will also find a list of common programs that
are known to work with @code{MyODBC}.


@node Installing MyODBC, ODBC administrator, ODBC, ODBC
@subsection How To Install MyODBC

@code{MyODBC} is a 32-bit ODBC (2.50) level 0 (with level 1 and
level 2 features) driver for connecting an ODBC-aware application
to MySQL.  @code{MyODBC} works on Windows 9x/Me/NT/2000/XP
and most Unix platforms.

@code{MyODBC} is in public domain, and you can find the newest
version at @uref{http://www.mysql.com/downloads/api-myodbc.html}.

If you have problem with @code{MyODBC} and your program also works
with OLEDB, you should try the OLEDB driver.

Normally you only need to install @code{MyODBC} on Windows machines.
You only need @code{MyODBC} for Unix if you have a program like
ColdFusion that is running on the Unix machine and uses ODBC to connect
to the databases.

If you want to install @code{MyODBC} on a Unix box, you will also need
an @code{ODBC} manager.  @code{MyODBC} is known to work with
most of the Unix ODBC managers.  @xref{Portals}.

To install @code{MyODBC} on Windows, you should download the
appropriate @code{MyODBC} @file{.zip} file,
unpack it with @code{WinZIP} or some similar program,
and execute the @file{SETUP.EXE} file.

On Windows/NT/XP you may get the following error when trying to install
@code{MyODBC}:

@example
An error occurred while copying C:\WINDOWS\SYSTEM\MFC30.DLL. Restart
Windows and try installing again (before running any applications which
use ODBC)
@end example

The problem in this case is that some other program is using ODBC and
because of how Windows is designed, you may not in this case be able to
install a new ODBC drivers with Microsoft's ODBC setup program. In most
cases you can continue by just pressing @code{Ignore} to copy the rest
of the MyODBC files and the final installation should still work.  If
this doesn't work, the solution is to reboot your computer in ``safe
mode`` (Choose this by pressing F8 just before your machine starts
Windows during rebooting), install @code{MyODBC}, and reboot to normal
mode.

@itemize @bullet
@item
To make a connection to a Unix box from a Windows box, with an ODBC
application (one that doesn't support MySQL natively), you must
first install @code{MyODBC} on the Windows machine.
@item
The user and Windows machine must have the access privileges to the
MySQL server on the Unix machine.  This is set up with the
@code{GRANT} command. @xref{GRANT,,@code{GRANT}}.
@item
You must create an ODBC DSN entry as follows:

@itemize @minus
@item
Open the Control Panel on the Windows machine.
@item
Double-click the ODBC Data Sources 32-bit icon.
@item
Click the tab User DSN.
@item
Click the button Add.
@item
Select MySQL in the screen Create New Data Source and click
the Finish button.
@item
The MySQL Driver default configuration screen is shown.
@xref{ODBC administrator}.
@end itemize

@item
Now start your application and select the ODBC driver with the DSN you
specified in the ODBC administrator.
@end itemize

Notice that there are other configuration options on the screen of
MySQL (trace, don't prompt on connect, etc) that you can try if
you run into problems.


@node ODBC administrator, MyODBC connect parameters, Installing MyODBC, ODBC
@subsection How to Fill in the Various Fields in the ODBC Administrator Program

@cindex ODBC, administrator

There are three possibilities for specifying the server name on
Windows95:

@itemize @bullet
@item
Use the IP address of the server.
@item
Add a file @file{\windows\lmhosts} with the following information:

@example
ip hostname
@end example

For example:

@example
194.216.84.21 my_hostname
@end example

@item
Configure the PC to use DNS.
@end itemize

Example of how to fill in the @code{ODBC setup}:
@example
Windows DSN name:   test
Description:        This is my test database
MySql Database:     test
Server:             194.216.84.21
User:               monty
Password:           my_password
Port:
@end example

The value for the @code{Windows DSN name} field is any name that is unique
in your Windows ODBC setup.

You don't have to specify values for the @code{Server}, @code{User},
@code{Password}, or @code{Port} fields in the ODBC setup screen.
However, if you do, the values will be used as the defaults later when
you attempt to make a connection.  You have the option of changing the
values at that time.

If the port number is not given, the default port (@value{default_port})
is used.

If you specify the option @code{Read options from C:\my.cnf}, the groups
@code{client} and @code{odbc} will be read from the @file{C:\my.cnf} file.
You can use all options that are usable by @code{mysql_options()}.
@xref{mysql_options, , @code{mysql_options()}}.


@node MyODBC connect parameters, ODBC Problems, ODBC administrator, ODBC
@subsection Connect parameters for MyODBC

One can specify the following parameters for @code{MyODBC} on
the @code{[Servername]} section of an @file{ODBC.INI} file or
through  the @code{InConnectionString} argument in the
@code{SQLDriverConnect()} call.

@multitable @columnfractions .15 .20 .65
@item @strong{Parameter} @tab @strong{Default value} @tab @strong{Comment}
@item user @tab ODBC (on Windows) @tab The username used to connect to MySQL.
@item server @tab localhost @tab The hostname of the MySQL server.
@item database @tab @tab The default database
@item option @tab 0 @tab A integer by which you can specify how @code{MyODBC} should work. See below.
@item port @tab 3306 @tab The TCP/IP port to use if @code{server} is not @code{localhost}.
@item stmt @tab @tab A statement that will be executed when connection to @code{MySQL}.
@item password @tab @tab The password for the @code{server} @code{user} combination.
@item socket @tab @tab The socket or Windows pipe to connect to.
@end multitable

The option argument is used to tell @code{MyODBC} that the client isn't 100%
ODBC compliant. On Windows, one normally sets the option flag by
toggling the different options on the connection screen but one can also
set this in the opton argument. The following options are listed in the
same order as they appear in the @code{MyODBC} connect screen:

@multitable @columnfractions .10 .90
@item @strong{Bit} @tab @strong{Description}
@item 1 @tab The client can't handle that @code{MyODBC} returns the real width of a column.
@item 2 @tab  The client can't handle that MySQL returns the true value of affected rows.  If this flag is set then MySQL returns 'found rows' instead.  One must have MySQL 3.21.14 or newer to get this to work.
@item 4 @tab Make a debug log in c:\myodbc.log.  This is the same as putting @code{MYSQL_DEBUG=d:t:O,c::\myodbc.log} in @file{AUTOEXEC.BAT}
@item 8  @tab Don't set any packet limit for results and parameters.
@item 16 @tab Don't prompt for questions even if driver would like to prompt
@item 32 @tab Simulate a ODBC 1.0 driver in some context.
@item 64 @tab Ignore use of database name in 'database.table.column'.
@item 128 @tab Force use of ODBC manager cursors (experimental).
@item 256 @tab Disable the use of extended fetch (experimental).
@item 512 @tab Pad CHAR fields to full column length.
@item 1024 @tab SQLDescribeCol() will return fully qualifed column names
@item 2048 @tab Use the compressed server/client protocol
@item 4096 @tab Tell server to ignore space after function name and before @code{'('} (needed by PowerBuilder).  This will make all function names keywords!
@item 8192 @tab Connect with named pipes to a @code{mysqld} server running on NT.
@item 16384 @tab Change LONGLONG columns to INT columns (some applications can't handle LONGLONG).
@item 32768 @tab Return 'user' as Table_qualifier and Table_owner from SQLTables (experimental)
@item 65536 @tab Read parameters from the @code{client} and @code{odbc} groups from @file{my.cnf}
@item 131072 @tab Add some extra safety checks (should not bee needed but...)
@end multitable

If you want to have many options, you should add the above flags!  For
example setting option to 12 (4+8) gives you debugging without package
limits!

The default @file{MYODBC.DLL} is compiled for optimal performance.  If
you want to to debug @code{MyODBC} (for example to enable tracing),
you should instead use @file{MYODBCD.DLL}.  To install this file, copy
@file{MYODBCD.DLL} over the installed @file{MYODBC.DLL} file.


@node ODBC Problems, MyODBC clients, MyODBC connect parameters, ODBC
@subsection How to Report Problems with MyODBC

@code{MyODBC} has been tested with Access, Admndemo.exe, C++-Builder,
Borland Builder 4, Centura Team Developer (formerly Gupta SQL/Windows),
ColdFusion (on Solaris and NT with svc pack 5), Crystal Reports,
DataJunction, Delphi, ERwin, Excel, iHTML, FileMaker Pro, FoxPro, Notes
4.5/4.6, SBSS, Perl DBD-ODBC, Paradox, Powerbuilder, Powerdesigner 32
bit, VC++, and Visual Basic.

If you know of any other applications that work with @code{MyODBC}, please
send mail to @email{myodbc@@lists.mysql.com} about this!

With some programs you may get an error like:
@code{Another user has modifies the record that you have modified}. In most
cases this can be solved by doing one of the following things:

@itemize @bullet
@item
Add a primary key for the table if there isn't one already.
@item
Add a timestamp column if there isn't one already.
@item
Only use double float fields. Some programs may fail when they compare
single floats.
@end itemize

If the above doesn't help, you should do a @code{MyODBC} trace file and
try to figure out why things go wrong.


@node MyODBC clients, ODBC and last_insert_id, ODBC Problems, ODBC
@subsection Programs Known to Work with MyODBC

Most programs should work with @code{MyODBC}, but for each of those
listed here, we have tested it ourselves or received confirmation from
some user that it works:

@table @asis
@item @strong{Program}
@strong{Comment}
@cindex Access program

@item Access
To make Access work:
@itemize @bullet
@item
If you are using Access 2000, you should get and install the newest
(version 2.6 or above) Microsoft MDAC (@code{Microsoft Data Access
Components}) from @uref{http://www.microsoft.com/data/}.  This will fix
the following bug in Access: when you export data to MySQL, the
table and column names aren't specified.  Another way to around this bug
is to upgrade to MyODBC Version 2.50.33 and MySQL Version
3.23.x, which together provide a workaround for this bug!

You should also get and apply the Microsoft Jet 4.0 Service Pack 5 (SP5)
which can be found here
@uref{http://support.microsoft.com/support/kb/articles/Q 239/1/14.ASP}.
This will fix some cases where columns are marked as @code{#deleted#}
in Access.

Note that if you are using MySQL Version 3.22, you must to apply the
MDAC patch and use MyODBC 2.50.32 or 2.50.34 and above to go around
this problem.
@item
For all Access versions, you should enable the MyODBC option flag
@code{Return matching rows}. For Access 2.0, you should additionally enable
@code{Simulate ODBC 1.0}.
@item
You should have a timestamp in all tables you want to be able to update.
For maximum portability @code{TIMESTAMP(14)} or simple @code{TIMESTAMP}
is recommended instead of other @code{TIMESTAMP(X)} variations.
@item
You should have a primary key in the table. If not, new or updated rows
may show up as @code{#DELETED#}.
@item
Only use @code{DOUBLE} float fields. Access fails when comparing with
single floats.  The symptom usually is that new or updated rows may show
up as @code{#DELETED#} or that you can't find or update rows.
@item
If you are linking a table through MyODBC, which has @code{BIGINT} as
one of the column, then the results will be displayed as @code{#DELETED}. The
work around solution is:
@itemize @bullet
@item
Have one more dummy column with @code{TIMESTAMP} as the data type, preferably
@code{TIMESTAMP(14)}.
@item
Check the @code{'Change BIGINT columns to INT'} in connection options dialog in
ODBC DSN Administrator
@item
Delete the table link from access and re-create it.
@end itemize

It still displays the previous records as @code{#DELETED#}, but newly
added/updated records will be displayed properly.
@item
If you still get the error @code{Another user has changed your data} after
adding a @code{TIMESTAMP} column, the following trick may help you:

Don't use @code{table} data sheet view. Create instead a form with the
fields you want, and use that @code{form} data sheet view.  You should
set the @code{DefaultValue} property for the @code{TIMESTAMP} column to
@code{NOW()}.  It may be a good idea to hide the @code{TIMESTAMP} column
from view so your users are not confused.
@item
In some cases, Access may generate illegal SQL queries that
MySQL can't understand. You can fix this by selecting
@code{"Query|SQLSpecific|Pass-Through"} from the Access menu.
@item
Access on NT will report @code{BLOB} columns as @code{OLE OBJECTS}. If
you want to have @code{MEMO} columns instead, you should change the
column to @code{TEXT} with @code{ALTER TABLE}.
@item
Access can't always handle @code{DATE} columns properly. If you have a problem
with these, change the columns to @code{DATETIME}.
@item
If you have in Access a column defined as @code{BYTE}, Access will try
to export this as @code{TINYINT} instead of @code{TINYINT UNSIGNED}.
This will give you problems if you have values > 127 in the column!
@end itemize

@cindex ADO program
@item ADO
When you are coding with the ADO API and @code{MyODBC} you need to put
attention in some default properties that aren't supported by the
MySQL server.  For example, using the @code{CursorLocation
Property} as @code{adUseServer} will return for the @code{RecordCount
Property} a result of -1. To have the right value, you need to set this
property to @code{adUseClient}, like is showing in the VB code here:

@example
Dim myconn As New ADODB.Connection
Dim myrs As New Recordset
Dim mySQL As String
Dim myrows As Long

myconn.Open "DSN=MyODBCsample"
mySQL = "SELECT * from user"
myrs.Source = mySQL
Set myrs.ActiveConnection = myconn
myrs.CursorLocation = adUseClient
myrs.Open
myrows = myrs.RecordCount

myrs.Close
myconn.Close
@end example

Another workaround is to use a @code{SELECT COUNT(*)} statement
for a similar query to get the correct row count.

@item Active server pages (ASP)
You should use the option flag @code{Return matching rows}.

@item BDE applications
To get these to work, you should set the option flags
@code{Don't optimize column widths} and @code{Return matching rows}.

@cindex Borland Builder 4 program
@item Borland Builder 4
When you start a query you can use the property @code{Active} or use the
method @code{Open}.  Note that @code{Active} will start by automatically
issuing a @code{SELECT * FROM ...} query that may not be a good thing if
your tables are big!
@item ColdFusion (On Unix)
The following information is taken from the ColdFusion documentation:

Use the following information to configure ColdFusion Server for Linux
to use the unixODBC driver with @code{MyODBC} for MySQL data
sources.  Allaire has verified that @code{MyODBC} Version 2.50.26
works with MySQL Version 3.22.27 and ColdFusion for Linux. (Any
newer version should also work.) You can download @code{MyODBC} at
@uref{http://www.mysql.com/downloads/api-myodbc.html}

@cindex ColdFusion program
ColdFusion Version 4.5.1 allows you to us the ColdFusion Administrator
to add the MySQL data source. However, the driver is not
included with ColdFusion Version 4.5.1. Before the MySQL driver
will appear in the ODBC datasources drop-down list, you must build and
copy the @code{MyODBC} driver to
@file{/opt/coldfusion/lib/libmyodbc.so}.

The Contrib directory contains the program @file{mydsn-xxx.zip} which allows
you to build and remove the DSN registry file for the MyODBC driver
on Coldfusion applications.

@cindex DataJunction
@item DataJunction
You have to change it to output @code{VARCHAR} rather than @code{ENUM}, as
it exports the latter in a manner that causes MySQL grief.
@cindex Excel
@item Excel
Works. A few tips:
@itemize @bullet
@item
If you have problems with dates, try to select them as strings using the
@code{CONCAT()} function. For example:
@example
select CONCAT(rise_time), CONCAT(set_time)
    from sunrise_sunset;
@end example
Values retrieved as strings this way should be correctly recognised
as time values by Excel97.

The purpose of @code{CONCAT()} in this example is to fool ODBC into thinking
the column is of ``string type''. Without the @code{CONCAT()}, ODBC knows the
column is of time type, and Excel does not understand that.

Note that this is a bug in Excel, because it automatically converts a
string to a time. This would be great if the source was a text file, but
is plain stupid when the source is an ODBC connection that reports
exact types for each column.
@end itemize
@cindex Word program
@item Word

To retrieve data from MySQL to Word/Excel documents, you need to
use the @code{MyODBC} driver and the Add-in Microsoft Query help.

For example, create a db with a table containing 2 columns of text:

@itemize @bullet
@item
Insert rows using the @code{mysql} client command-line tool.
@item
Create a DSN file using the ODBC manager, for example, @file{my} for the db above.
@item
Open the Word application.
@item
Create a blank new documentation.
@item
Using the tool bar called Database, press the button insert database.
@item
Press the button Get Data.
@item
At the right hand of the screen Get Data, press the button Ms Query.
@item
In the Ms Query create a New Data Source using the DSN file my.
@item
Select the new query.
@item
Select the columns that you want.
@item
Make a filter if you want.
@item
Make a Sort if you want.
@item
Select Return Data to Microsoft Word.
@item
Click Finish.
@item
Click Insert data and select the records.
@item
Click OK and you see the rows in your Word document.
@end itemize

@cindex odbcadmin program
@item odbcadmin
Test program for ODBC.
@cindex Delphi program
@item Delphi
You must use BDE Version 3.2 or newer.  Set the @code{Don't optimize column width}
option field when connecting to MySQL.

Also, here is some potentially useful Delphi code that sets up both an
ODBC entry and a BDE entry for @code{MyODBC} (the BDE entry requires a BDE
Alias Editor that is free at a Delphi Super Page near
you. (Thanks to Bryan Brunton @email{bryan@@flesherfab.com} for this):

@example
fReg:= TRegistry.Create;
  fReg.OpenKey('\Software\ODBC\ODBC.INI\DocumentsFab', True);
  fReg.WriteString('Database', 'Documents');
  fReg.WriteString('Description', ' ');
  fReg.WriteString('Driver', 'C:\WINNT\System32\myodbc.dll');
  fReg.WriteString('Flag', '1');
  fReg.WriteString('Password', '');
  fReg.WriteString('Port', ' ');
  fReg.WriteString('Server', 'xmark');
  fReg.WriteString('User', 'winuser');
  fReg.OpenKey('\Software\ODBC\ODBC.INI\ODBC Data Sources', True);
  fReg.WriteString('DocumentsFab', 'MySQL');
  fReg.CloseKey;
  fReg.Free;

  Memo1.Lines.Add('DATABASE NAME=');
  Memo1.Lines.Add('USER NAME=');
  Memo1.Lines.Add('ODBC DSN=DocumentsFab');
  Memo1.Lines.Add('OPEN MODE=READ/WRITE');
  Memo1.Lines.Add('BATCH COUNT=200');
  Memo1.Lines.Add('LANGDRIVER=');
  Memo1.Lines.Add('MAX ROWS=-1');
  Memo1.Lines.Add('SCHEMA CACHE DIR=');
  Memo1.Lines.Add('SCHEMA CACHE SIZE=8');
  Memo1.Lines.Add('SCHEMA CACHE TIME=-1');
  Memo1.Lines.Add('SQLPASSTHRU MODE=SHARED AUTOCOMMIT');
  Memo1.Lines.Add('SQLQRYMODE=');
  Memo1.Lines.Add('ENABLE SCHEMA CACHE=FALSE');
  Memo1.Lines.Add('ENABLE BCD=FALSE');
  Memo1.Lines.Add('ROWSET SIZE=20');
  Memo1.Lines.Add('BLOBS TO CACHE=64');
  Memo1.Lines.Add('BLOB SIZE=32');

  AliasEditor.Add('DocumentsFab','MySQL',Memo1.Lines);
@end example

@cindex C++ Builder
@item C++ Builder
Tested with BDE Version 3.0. The only known problem is that when the table
schema changes, query fields are not updated. BDE, however, does not seem
to recognise primary keys, only the index PRIMARY, though this has not
been a problem.

@item Vision
You should use the option flag @code{Return matching rows}.

@cindex Visual Basic
@item Visual Basic
To be able to update a table, you must define a primary key for the table.

Visual Basic with ADO can't handle big integers. This means that some queries
like @code{SHOW PROCESSLIST} will not work properly.  The fix is to set
add the option @code{OPTION=16834} in the ODBC connect string or set
the @code{Change BIGINT columns to INT} option in the MyODBC connect screen.
You may also want to set the @code{Return matching rows} option.

@item VisualInterDev
If you get the error @code{[Microsoft][ODBC Driver Manager] Driver does
not support this parameter} the reason may be that you have a
@code{BIGINT} in your result.  Try setting the @code{Change BIGINT
columns to INT} option in the MyODBC connect screen.

@item Visual Objects
You should use the option flag @code{Don't optimize column widths}.
@end table


@node ODBC and last_insert_id, MyODBC bug report, MyODBC clients, ODBC
@subsection How to Get the Value of an @code{AUTO_INCREMENT} Column in ODBC

@cindex AUTO-INCREMENT, ODBC

A common problem is how to get the value of an automatically generated ID
from an @code{INSERT}. With ODBC, you can do something like this (assuming
that @code{auto} is an @code{AUTO_INCREMENT} field):

@example
INSERT INTO foo (auto,text) VALUES(NULL,'text');
SELECT LAST_INSERT_ID();
@end example

Or, if you are just going to insert the ID into another table, you can do this:

@example
INSERT INTO foo (auto,text) VALUES(NULL,'text');
INSERT INTO foo2 (id,text) VALUES(LAST_INSERT_ID(),'text');
@end example

@xref{Getting unique ID}.

For the benefit of some ODBC applications (at least Delphi and Access),
the following query can be used to find a newly inserted row:
@example
SELECT * FROM tbl_name WHERE auto IS NULL;
@end example


@node MyODBC bug report,  , ODBC and last_insert_id, ODBC
@subsection Reporting Problems with MyODBC

@cindex reporting, MyODBC problems
@cindex problems, ODBC
@cindex MyODBC, reporting problems

If you encounter difficulties with @code{MyODBC}, you should start by
making a log file from the ODBC manager (the log you get when requesting
logs from ODBCADMIN) and a @code{MyODBC} log.

To get a @code{MyODBC} log, you need to do the following:

@enumerate
@item
Ensure that you are using @file{myodbcd.dll} and not @file{myodbc.dll}.
The easiest way to do this is to get @file{myodbcd.dll} from the MyODBC
distribution and copy it over the @file{myodbc.dll}, which is probably
in your @file{C:\windows\system32} or @file{C:\winnt\system32} directory.

Note that you probably want to restore the old myodbc.dll file when you
have finished testing, as this is a lot faster than @file{myodbcd.dll}.
@item
Tag the `Trace MyODBC' option flag in the @code{MyODBC} connect/configure
screen.  The log will be written to file @file{C:\myodbc.log}.

If the trace option is not remembered when you are going back to the
above screen, it means that you are not using the @code{myodbcd.dll}
driver (see the item above).
@item
Start your application and try to get it to fail.
@end enumerate

Check the @code{MyODBC trace file}, to find out what could be wrong.
You should be able to find out the issued queries by searching after
the string @code{>mysql_real_query} in the @file{myodbc.log} file.

You should also try duplicating the queries in the @code{mysql} monitor
or @code{admndemo} to find out if the error is MyODBC or MySQL.

If you find out something is wrong, please only send the relevant rows
(max 40 rows) to @email{myodbc@@lists.mysql.com}.  Please never
send the whole MyODBC or ODBC log file!

If you are unable to find out what's wrong, the last option is to
make an archive (tar or zip) that contains a MyODBC trace file, the ODBC
log file, and a README file that explains the problem.  You can send this
to @uref{ftp://support.mysql.com/pub/mysql/secret/}.  Only we at
MySQL AB will have access to the files you upload, and we will
be very discrete with the data!

If you can create a program that also shows this problem, please
upload this too!

If the program works with some other SQL server, you should make an ODBC log
file where you do exactly the same thing in the other SQL server.

Remember that the more information you can supply to us, the more
likely it is that we can fix the problem!


@node C, Cplusplus, ODBC, Clients
@section MySQL C API

@cindex C API, datatypes
@cindex datatypes, C API

@menu
* C API datatypes::             C API Datatypes
* C API function overview::     C API Function Overview
* C API functions::             C API Function Descriptions
* C Thread functions::          C Thread Functions
* C Embedded Server func::      C Embedded Server Function Descriptions
* C API problems::              Common questions and problems when using the C API
* Building clients::            Building Client Programs
* Threaded clients::            How to Make a Threaded Client
* libmysqld::                   libmysqld, the Embedded MySQL Server Library
@end menu

The C API code is distributed with MySQL. It is included in the
@code{mysqlclient} library and allows C programs to access a database.

Many of the clients in the MySQL source distribution are
written in C.  If you are looking for examples that demonstrate how to
use the C API, take a look at these clients. You can find these in the
@code{clients} directory in the MySQL source distribution.

Most of the other client APIs (all except Java) use the @code{mysqlclient}
library to communicate with the MySQL server. This means that, for
example, you can take advantage of many of the same environment variables
that are used by other client programs, because they are referenced from the
library.  See @ref{Client-Side Scripts}, for a list of these variables.

The client has a maximum communication buffer size. The size of the buffer
that is allocated initially (16K bytes) is automatically increased up to the
maximum size (the maximum is 16M). Because buffer sizes are increased
only as demand warrants, simply increasing the default maximum limit does not
in itself cause more resources to be used. This size check is mostly a check
for erroneous queries and communication packets.

The communication buffer must be large enough to contain a single SQL
statement (for client-to-server traffic) and one row of returned data (for
server-to-client traffic).  Each thread's communication buffer is dynamically
enlarged to handle any query or row up to the maximum limit.  For example, if
you have @code{BLOB} values that contain up to 16M of data, you must have a
communication buffer limit of at least 16M (in both server and client).  The
client's default maximum is 16M, but the default maximum in the server is
1M.  You can increase this by changing the value of the
@code{max_allowed_packet} parameter when the server is started.  @xref{Server
parameters}.

The MySQL server shrinks each communication buffer to
@code{net_buffer_length} bytes after each query.  For clients, the size of
the buffer associated with a connection is not decreased until the connection
is closed, at which time client memory is reclaimed.

For programming with threads, see @ref{Threaded clients}.
For creating a stand-alone application which includes the
"server" and "client" in the same program (and does not
communicate with an external MySQL server), see @ref{libmysqld}.


@node C API datatypes, C API function overview, C, C
@subsection C API Datatypes

@table @code
@tindex MYSQL C type
@item MYSQL
This structure represents a handle to one database connection. It is
used for almost all MySQL functions.

@tindex MYSQL_RES C type
@item MYSQL_RES
This structure represents the result of a query that returns rows
(@code{SELECT}, @code{SHOW}, @code{DESCRIBE}, @code{EXPLAIN}).  The
information returned from a query is called the @emph{result set} in the
remainder of this section.

@tindex MYSQL_ROW C type
@item MYSQL_ROW
This is a type-safe representation of one row of data. It is currently
implemented as an array of counted byte strings.  (You cannot treat these as
null-terminated strings if field values may contain binary data, because such
values may contain null bytes internally.)  Rows are obtained by calling
@code{mysql_fetch_row()}.

@tindex MYSQL_FIELD C type
@item MYSQL_FIELD
This structure contains information about a field, such as the field's
name, type, and size. Its members are described in more detail here.
You may obtain the @code{MYSQL_FIELD} structures for each field by
calling @code{mysql_fetch_field()} repeatedly.  Field values are not part of
this structure; they are contained in a @code{MYSQL_ROW} structure.


@tindex MYSQL_FIELD_OFFSET C type
@item MYSQL_FIELD_OFFSET
This is a type-safe representation of an offset into a MySQL field
list.  (Used by @code{mysql_field_seek()}.)  Offsets are field numbers
within a row, beginning at zero.

@tindex my_ulonglong C type
@tindex my_ulonglong values, printing
@item my_ulonglong
The type used for the number of rows and for @code{mysql_affected_rows()},
@code{mysql_num_rows()}, and @code{mysql_insert_id()}. This type provides a
range of @code{0} to @code{1.84e19}.

On some systems, attempting to print a value of type @code{my_ulonglong}
will not work.  To print such a value, convert it to @code{unsigned long}
and use a @code{%lu} print format.  Example:
@example
printf (Number of rows: %lu\n", (unsigned long) mysql_num_rows(result));
@end example
@end table

@noindent
The @code{MYSQL_FIELD} structure contains the members listed here:

@table @code
@item char * name
The name of the field, as a null-terminated string.

@item char * table
The name of the table containing this field, if it isn't a calculated field.
For calculated fields, the @code{table} value is an empty string.

@item char * def
The default value of this field, as a null-terminated string.  This is set
only if you use @code{mysql_list_fields()}.

@item enum enum_field_types type
The type of the field.
The @code{type} value may be one of the following:

@multitable @columnfractions .35 .55
@item @strong{Type value}                @tab @strong{Type description}
@item @code{FIELD_TYPE_TINY}            @tab @code{TINYINT} field
@item @code{FIELD_TYPE_SHORT}           @tab @code{SMALLINT} field
@item @code{FIELD_TYPE_LONG}            @tab @code{INTEGER} field
@item @code{FIELD_TYPE_INT24}           @tab @code{MEDIUMINT} field
@item @code{FIELD_TYPE_LONGLONG}        @tab @code{BIGINT} field
@item @code{FIELD_TYPE_DECIMAL}         @tab @code{DECIMAL} or @code{NUMERIC} field
@item @code{FIELD_TYPE_FLOAT}           @tab @code{FLOAT} field
@item @code{FIELD_TYPE_DOUBLE}          @tab @code{DOUBLE} or @code{REAL} field
@item @code{FIELD_TYPE_TIMESTAMP}       @tab @code{TIMESTAMP} field
@item @code{FIELD_TYPE_DATE}            @tab @code{DATE} field
@item @code{FIELD_TYPE_TIME}            @tab @code{TIME} field
@item @code{FIELD_TYPE_DATETIME}        @tab @code{DATETIME} field
@item @code{FIELD_TYPE_YEAR}            @tab @code{YEAR} field
@item @code{FIELD_TYPE_STRING}          @tab String (@code{CHAR} or @code{VARCHAR}) field
@item @code{FIELD_TYPE_BLOB}            @tab @code{BLOB} or @code{TEXT} field (use @code{max_length} to determine the maximum length)
@item @code{FIELD_TYPE_SET}             @tab @code{SET} field
@item @code{FIELD_TYPE_ENUM}            @tab @code{ENUM} field
@item @code{FIELD_TYPE_NULL}            @tab @code{NULL}-type field
@item @code{FIELD_TYPE_CHAR}            @tab Deprecated; use @code{FIELD_TYPE_TINY} instead
@end multitable

You can use the @code{IS_NUM()} macro to test whether a field has a
numeric type.  Pass the @code{type} value to @code{IS_NUM()} and it
will evaluate to TRUE if the field is numeric:

@example
if (IS_NUM(field->type))
    printf("Field is numeric\n");
@end example

@item unsigned int length
The width of the field, as specified in the table definition.

@item unsigned int max_length
The maximum width of the field for the result set (the length of the longest
field value for the rows actually in the result set). If you use
@code{mysql_store_result()} or @code{mysql_list_fields()}, this contains the
maximum length for the field.  If you use @code{mysql_use_result()}, the
value of this variable is zero.

@item unsigned int flags
Different bit-flags for the field.  The @code{flags} value may have zero
or more of the following bits set:

@multitable @columnfractions .35 .55
@item @strong{Flag value}         @tab @strong{Flag description}
@item @code{NOT_NULL_FLAG}       @tab Field can't be @code{NULL}
@item @code{PRI_KEY_FLAG}        @tab Field is part of a primary key
@item @code{UNIQUE_KEY_FLAG}     @tab Field is part of a unique key
@item @code{MULTIPLE_KEY_FLAG}   @tab Field is part of a non-unique key
@item @code{UNSIGNED_FLAG}       @tab Field has the @code{UNSIGNED} attribute
@item @code{ZEROFILL_FLAG}       @tab Field has the @code{ZEROFILL} attribute
@item @code{BINARY_FLAG}         @tab Field has the @code{BINARY} attribute
@item @code{AUTO_INCREMENT_FLAG} @tab Field has the @code{AUTO_INCREMENT}
attribute
@item @code{ENUM_FLAG}           @tab Field is an @code{ENUM} (deprecated)
@item @code{SET_FLAG}            @tab Field is a @code{SET} (deprecated)
@item @code{BLOB_FLAG}           @tab Field is a @code{BLOB} or @code{TEXT} (deprecated)
@item @code{TIMESTAMP_FLAG}      @tab Field is a @code{TIMESTAMP} (deprecated)
@end multitable

Use of the @code{BLOB_FLAG}, @code{ENUM_FLAG}, @code{SET_FLAG}, and
@code{TIMESTAMP_FLAG} flags is deprecated because they indicate the type of
a field rather than an attribute of its type.  It is preferable to test
@code{field->type} against @code{FIELD_TYPE_BLOB}, @code{FIELD_TYPE_ENUM},
@code{FIELD_TYPE_SET}, or @code{FIELD_TYPE_TIMESTAMP} instead.

@noindent
The following example illustrates a typical use of the @code{flags} value:

@example
if (field->flags & NOT_NULL_FLAG)
    printf("Field can't be null\n");
@end example

You may use the following convenience macros to determine the boolean
status of the @code{flags} value:

@multitable @columnfractions .25 .55
@item @strong{Flag status} @tab @strong{Description}
@item @code{IS_NOT_NULL(flags)} @tab True if this field is defined as @code{NOT NULL}
@item @code{IS_PRI_KEY(flags)}  @tab True if this field is a primary key
@item @code{IS_BLOB(flags)}     @tab True if this field is a @code{BLOB} or @code{TEXT} (deprecated; test @code{field->type} instead)
@end multitable

@item unsigned int decimals
The number of decimals for numeric fields.
@end table


@node C API function overview, C API functions, C API datatypes, C
@subsection C API Function Overview

@cindex C API, functions
@cindex functions, C API

The functions available in the C API are listed here and are described in
greater detail in a later section.
@xref{C API functions}.

@multitable @columnfractions .32 .68
@item @strong{Function} @tab @strong{Description}

@item @strong{mysql_affected_rows()} @tab
Returns the number of rows changed/deleted/inserted by the last @code{UPDATE},
@code{DELETE}, or @code{INSERT} query.

@item @strong{mysql_change_user()} @tab
Changes user and database on an open connection.

@item @strong{mysql_character_set_name()} @tab
Returns the name of the default character set for the connection.

@item @strong{mysql_close()} @tab
Closes a server connection.

@item @strong{mysql_connect()} @tab
Connects to a MySQL server. This function is deprecated; use
@code{mysql_real_connect()} instead.

@item @strong{mysql_create_db()} @tab
Creates a database. This function is deprecated; use the SQL command
@code{CREATE DATABASE} instead.

@item @strong{mysql_data_seek()} @tab
Seeks to an arbitrary row in a query result set.

@item @strong{mysql_debug()} @tab
Does a @code{DBUG_PUSH} with the given string.

@item @strong{mysql_drop_db()} @tab
Drops a database. This function is deprecated; use the SQL command
@code{DROP DATABASE} instead.

@item @strong{mysql_dump_debug_info()} @tab
Makes the server write debug information to the log.

@item @strong{mysql_eof()} @tab
Determines whether the last row of a result set has been read.
This function is deprecated; @code{mysql_errno()} or @code{mysql_error()}
may be used instead.

@item @strong{mysql_errno()} @tab
Returns the error number for the most recently invoked MySQL function.

@item @strong{mysql_error()} @tab
Returns the error message for the most recently invoked MySQL function.

@item @strong{mysql_escape_string()} @tab
Escapes special characters in a string for use in a SQL statement.

@item @strong{mysql_fetch_field()} @tab
Returns the type of the next table field.

@item @strong{mysql_fetch_field_direct()} @tab
Returns the type of a table field, given a field number.

@item @strong{mysql_fetch_fields()} @tab
Returns an array of all field structures.

@item @strong{mysql_fetch_lengths()} @tab
Returns the lengths of all columns in the current row.

@item @strong{mysql_fetch_row()} @tab
Fetches the next row from the result set.

@item @strong{mysql_field_seek()} @tab
Puts the column cursor on a specified column.

@item @strong{mysql_field_count()} @tab
Returns the number of result columns for the most recent query.

@item @strong{mysql_field_tell()} @tab
Returns the position of the field cursor used for the last
@code{mysql_fetch_field()}.

@item @strong{mysql_free_result()} @tab
Frees memory used by a result set.

@item @strong{mysql_get_client_info()} @tab
Returns client version information.

@item @strong{mysql_get_host_info()} @tab
Returns a string describing the connection.

@item @strong{mysql_get_proto_info()} @tab
Returns the protocol version used by the connection.

@item @strong{mysql_get_server_info()} @tab
Returns the server version number.

@item @strong{mysql_info()} @tab
Returns information about the most recently executed query.

@item @strong{mysql_init()} @tab
Gets or initialises a @code{MYSQL} structure.

@item @strong{mysql_insert_id()} @tab
Returns the ID generated for an @code{AUTO_INCREMENT} column by the previous
query.

@item @strong{mysql_kill()} @tab
Kills a given thread.

@item @strong{mysql_list_dbs()} @tab
Returns database names matching a simple regular expression.

@item @strong{mysql_list_fields()} @tab
Returns field names matching a simple regular expression.

@item @strong{mysql_list_processes()} @tab
Returns a list of the current server threads.

@item @strong{mysql_list_tables()} @tab
Returns table names matching a simple regular expression.

@item @strong{mysql_num_fields()} @tab
Returns the number of columns in a result set.

@item @strong{mysql_num_rows()} @tab
Returns the number of rows in a result set.

@item @strong{mysql_options()} @tab
Sets connect options for @code{mysql_connect()}.

@item @strong{mysql_ping()} @tab
Checks whether the connection to the server is working, reconnecting
as necessary.

@item @strong{mysql_query()} @tab
Executes a SQL query specified as a null-terminated string.

@item @strong{mysql_real_connect()} @tab
Connects to a MySQL server.

@item @strong{mysql_real_escape_string()} @tab
Escapes special characters in a string for use in a SQL statement, taking
into account the current charset of the connection.

@item @strong{mysql_real_query()} @tab
Executes a SQL query specified as a counted string.

@item @strong{mysql_reload()} @tab
Tells the server to reload the grant tables.

@item @strong{mysql_row_seek()} @tab
Seeks to a row in a result set, using value returned from
@code{mysql_row_tell()}.

@item @strong{mysql_row_tell()} @tab
Returns the row cursor position.

@item @strong{mysql_select_db()} @tab
Selects a database.

@item @strong{mysql_shutdown()} @tab
Shuts down the database server.

@item @strong{mysql_stat()} @tab
Returns the server status as a string.

@item @strong{mysql_store_result()} @tab
Retrieves a complete result set to the client.

@item @strong{mysql_thread_id()} @tab
Returns the current thread ID.

@item @strong{mysql_thread_safe()} @tab
Returns 1 if the clients are compiled as thread-safe.

@item @strong{mysql_use_result()} @tab
Initiates a row-by-row result set retrieval.
@end multitable

To connect to the server, call @code{mysql_init()} to initialise a
connection handler, then call @code{mysql_real_connect()} with that
handler (along with other information such as the hostname, user name,
and password).  Upon connection, @code{mysql_real_connect()} sets the
@code{reconnect} flag (part of the MYSQL structure) to a value of
@code{1}. This flag indicates, in the event that a query cannot be
performed because of a lost connection, to try reconnecting to the
server before giving up.  When you are done with the connection, call
@code{mysql_close()} to terminate it.

While a connection is active, the client may send SQL queries to the server
using @code{mysql_query()} or @code{mysql_real_query()}.  The difference
between the two is that @code{mysql_query()} expects the query to be
specified as a null-terminated string whereas @code{mysql_real_query()}
expects a counted string.  If the string contains binary data (which may
include null bytes), you must use @code{mysql_real_query()}.

For each non-@code{SELECT} query (for example, @code{INSERT}, @code{UPDATE},
@code{DELETE}), you can find out how many rows were changed (affected)
by calling @code{mysql_affected_rows()}.

For @code{SELECT} queries, you retrieve the selected rows as a result set.
(Note that some statements are @code{SELECT}-like in that they return rows.
These include @code{SHOW}, @code{DESCRIBE}, and @code{EXPLAIN}.  They should
be treated the same way as @code{SELECT} statements.)

There are two ways for a client to process result sets.  One way is to
retrieve the entire result set all at once by calling
@code{mysql_store_result()}.  This function acquires from the server all the
rows returned by the query and stores them in the client.  The second way is
for the client to initiate a row-by-row result set retrieval by calling
@code{mysql_use_result()}.  This function initialises the retrieval, but does
not actually get any rows from the server.

In both cases, you access rows by calling @code{mysql_fetch_row()}.  With
@code{mysql_store_result()}, @code{mysql_fetch_row()} accesses rows that have
already been fetched from the server.  With @code{mysql_use_result()},
@code{mysql_fetch_row()} actually retrieves the row from the server.
Information about the size of the data in each row is available by
calling @code{mysql_fetch_lengths()}.

After you are done with a result set, call @code{mysql_free_result()}
to free the memory used for it.

The two retrieval mechanisms are complementary.  Client programs should
choose the approach that is most appropriate for their requirements.
In practice, clients tend to use @code{mysql_store_result()} more
commonly.

An advantage of @code{mysql_store_result()} is that because the rows have all
been fetched to the client, you not only can access rows sequentially, you
can move back and forth in the result set using @code{mysql_data_seek()} or
@code{mysql_row_seek()} to change the current row position within the result
set.  You can also find out how many rows there are by calling
@code{mysql_num_rows()}.  On the other hand, the memory requirements for
@code{mysql_store_result()} may be very high for large result sets and you
are more likely to encounter out-of-memory conditions.

An advantage of @code{mysql_use_result()} is that the client requires less
memory for the result set because it maintains only one row at a time (and
because there is less allocation overhead, @code{mysql_use_result()} can be
faster).  Disadvantages are that you must process each row quickly to avoid
tying up the server, you don't have random access to rows within the result
set (you can only access rows sequentially), and you don't know how many rows
are in the result set until you have retrieved them all.  Furthermore, you
@strong{must} retrieve all the rows even if you determine in mid-retrieval that
you've found the information you were looking for.

The API makes it possible for clients to respond appropriately to
queries (retrieving rows only as necessary) without knowing whether or
not the query is a @code{SELECT}.  You can do this by calling
@code{mysql_store_result()} after each @code{mysql_query()} (or
@code{mysql_real_query()}).  If the result set call succeeds, the query
was a @code{SELECT} and you can read the rows.  If the result set call
fails, call @code{mysql_field_count()} to determine whether a
result was actually to be expected.  If @code{mysql_field_count()}
returns zero, the query returned no data (indicating that it was an
@code{INSERT}, @code{UPDATE}, @code{DELETE}, etc.), and was not
expected to return rows.  If @code{mysql_field_count()} is non-zero, the
query should have returned rows, but didn't.  This indicates that the
query was a @code{SELECT} that failed.  See the description for
@code{mysql_field_count()} for an example of how this can be done.

Both @code{mysql_store_result()} and @code{mysql_use_result()} allow you to
obtain information about the fields that make up the result set (the number
of fields, their names and types, etc.).  You can access field information
sequentially within the row by calling @code{mysql_fetch_field()} repeatedly,
or by field number within the row by calling
@code{mysql_fetch_field_direct()}.  The current field cursor position may be
changed by calling @code{mysql_field_seek()}.  Setting the field cursor
affects subsequent calls to @code{mysql_fetch_field()}.  You can also get
information for fields all at once by calling @code{mysql_fetch_fields()}.

For detecting and reporting errors, MySQL provides access to error
information by means of the @code{mysql_errno()} and @code{mysql_error()}
functions.  These return the error code or error message for the most
recently invoked function that can succeed or fail, allowing you to determine
when an error occurred and what it was.


@node C API functions, C Thread functions, C API function overview, C
@subsection C API Function Descriptions

@menu
* mysql_affected_rows::         @code{mysql_affected_rows()}
* mysql_change_user::           @code{mysql_change_user()}
* mysql_character_set_name::    @code{mysql_character_set_name()}
* mysql_close::                 @code{mysql_close()}
* mysql_connect::               @code{mysql_connect()}
* mysql_create_db::             @code{mysql_create_db()}
* mysql_data_seek::             @code{mysql_data_seek()}
* mysql_debug::                 @code{mysql_debug()}
* mysql_drop_db::               @code{mysql_drop_db()}
* mysql_dump_debug_info::       @code{mysql_dump_debug_info()}
* mysql_eof::                   @code{mysql_eof()}
* mysql_errno::                 @code{mysql_errno()}
* mysql_error::                 @code{mysql_error()}
* mysql_escape_string::         @code{mysql_escape_string()}
* mysql_fetch_field::           @code{mysql_fetch_field()}
* mysql_fetch_fields::          @code{mysql_fetch_fields()}
* mysql_fetch_field_direct::    @code{mysql_fetch_field_direct()}
* mysql_fetch_lengths::         @code{mysql_fetch_lengths()}
* mysql_fetch_row::             @code{mysql_fetch_row()}
* mysql_field_count::           @code{mysql_field_count()}
* mysql_field_seek::            @code{mysql_field_seek()}
* mysql_field_tell::            @code{mysql_field_tell()}
* mysql_free_result::           @code{mysql_free_result()}
* mysql_get_client_info::       @code{mysql_get_client_info()}
* mysql_get_host_info::         @code{mysql_get_host_info()}
* mysql_get_proto_info::        @code{mysql_get_proto_info()}
* mysql_get_server_info::       @code{mysql_get_server_info()}
* mysql_info::                  @code{mysql_info()}
* mysql_init::                  @code{mysql_init()}
* mysql_insert_id::             @code{mysql_insert_id()}
* mysql_kill::                  @code{mysql_kill()}
* mysql_list_dbs::              @code{mysql_list_dbs()}
* mysql_list_fields::           @code{mysql_list_fields()}
* mysql_list_processes::        @code{mysql_list_processes()}
* mysql_list_tables::           @code{mysql_list_tables()}
* mysql_num_fields::            @code{mysql_num_fields()}
* mysql_num_rows::              @code{mysql_num_rows()}
* mysql_options::               @code{mysql_options()}
* mysql_ping::                  @code{mysql_ping()}
* mysql_query::                 @code{mysql_query()}
* mysql_real_connect::          @code{mysql_real_connect()}
* mysql_real_escape_string::    @code{mysql_real_escape_string()}
* mysql_real_query::            @code{mysql_real_query()}
* mysql_reload::                @code{mysql_reload()}
* mysql_row_seek::              @code{mysql_row_seek()}
* mysql_row_tell::              @code{mysql_row_tell()}
* mysql_select_db::             @code{mysql_select_db()}
* mysql_shutdown::              @code{mysql_shutdown()}
* mysql_stat::                  @code{mysql_stat()}
* mysql_store_result::          @code{mysql_store_result()}
* mysql_thread_id::             @code{mysql_thread_id()}
* mysql_use_result::            @code{mysql_use_result()}
@end menu

In the descriptions here, a parameter or return value of @code{NULL} means
@code{NULL} in the sense of the C programming language, not a
MySQL @code{NULL} value.

Functions that return a value generally return a pointer or an integer.
Unless specified otherwise, functions returning a pointer return a
non-@code{NULL} value to indicate success or a @code{NULL} value to indicate
an error, and functions returning an integer return zero to indicate success
or non-zero to indicate an error.  Note that ``non-zero'' means just that.
Unless the function description says otherwise, do not test against a value
other than zero:

@example
if (result)                   /* correct */
    ... error ...

if (result < 0)               /* incorrect */
    ... error ...

if (result == -1)             /* incorrect */
    ... error ...
@end example

When a function returns an error, the @strong{Errors} subsection of the
function description lists the possible types of errors.  You can
find out which of these occurred by calling @code{mysql_errno()}.
A string representation of the error may be obtained by calling
@code{mysql_error()}.


@node mysql_affected_rows, mysql_change_user, C API functions, C API functions
@subsubsection @code{mysql_affected_rows()}

@findex @code{mysql_affected_rows()}

@code{my_ulonglong mysql_affected_rows(MYSQL *mysql)}

@subsubheading Description

Returns the number of rows changed by the last @code{UPDATE}, deleted by
the last @code{DELETE} or inserted by the last @code{INSERT}
statement. May be called immediately after @code{mysql_query()} for
@code{UPDATE}, @code{DELETE}, or @code{INSERT} statements.  For
@code{SELECT} statements, @code{mysql_affected_rows()} works like
@code{mysql_num_rows()}.

@subsubheading Return Values

An integer greater than zero indicates the number of rows affected or
retrieved.  Zero indicates that no records where updated for an
@code{UPDATE} statement, no rows matched the @code{WHERE} clause in the
query or that no query has yet been executed.  -1 indicates that the
query returned an error or that, for a @code{SELECT} query,
@code{mysql_affected_rows()} was called prior to calling
@code{mysql_store_result()}.

@subsubheading Errors

None.

@subsubheading Example

@example
mysql_query(&mysql,"UPDATE products SET cost=cost*1.25 WHERE group=10");
printf("%ld products updated",(long) mysql_affected_rows(&mysql));
@end example

If one specifies the flag @code{CLIENT_FOUND_ROWS} when connecting to
@code{mysqld}, @code{mysql_affected_rows()} will return the number of
rows matched by the @code{WHERE} statement for @code{UPDATE} statements.

Note that when one uses a @code{REPLACE} command,
@code{mysql_affected_rows()} will return 2 if the new row replaced and
old row.  This is because in this case one row was inserted and then the
duplicate was deleted.


@node mysql_change_user, mysql_character_set_name, mysql_affected_rows, C API functions
@subsubsection @code{mysql_change_user()}

@findex @code{mysql_change_user()}

@code{my_bool mysql_change_user(MYSQL *mysql, const char *user, const
char *password, const char *db)}

@subsubheading Description

Changes the user and causes the database specified by @code{db} to
become the default (current) database on the connection specified by
@code{mysql}.  In subsequent queries, this database is the default for
table references that do not include an explicit database specifier.

This function was introduced in MySQL Version 3.23.3.

@code{mysql_change_user()} fails unless the connected user can be
authenticated or if he doesn't have permission to use the database.  In
this case the user and database are not changed

The @code{db} parameter may be set to @code{NULL} if you don't want to have a
default database.

@subsubheading Return Values

Zero for success.  Non-zero if an error occurred.

@subsubheading Errors

The same that you can get from @code{mysql_real_connect()}.

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@item ER_UNKNOWN_COM_ERROR
The MySQL server doesn't implement this command (probably an old server)
@item ER_ACCESS_DENIED_ERROR
The user or password was wrong.
@item ER_BAD_DB_ERROR
The database didn't exist.
@item ER_DBACCESS_DENIED_ERROR
The user did not have access rights to the database.
@item ER_WRONG_DB_NAME
The database name was too long.
@end table

@subsubheading Example

@example
if (mysql_change_user(&mysql, "user", "password", "new_database"))
@{
   fprintf(stderr, "Failed to change user.  Error: %s\n",
           mysql_error(&mysql));
@}
@end example


@node mysql_character_set_name, mysql_close, mysql_change_user, C API functions
@subsubsection @code{mysql_character_set_name()}

@findex @code{mysql_character_set_name()}

@code{const char *mysql_character_set_name(MYSQL *mysql)}

@subsubheading Description

Returns the default character set for the current connection.

@subsubheading Return Values

The default character set

@subsubheading Errors
None.


@node mysql_close, mysql_connect, mysql_character_set_name, C API functions
@subsubsection @code{mysql_close()}

@findex @code{mysql_close()}

@code{void mysql_close(MYSQL *mysql)}

@subsubheading Description
Closes a previously opened connection.  @code{mysql_close()} also deallocates
the connection handle pointed to by @code{mysql} if the handle was allocated
automatically by @code{mysql_init()} or @code{mysql_connect()}.

@subsubheading Return Values

None.

@subsubheading Errors

None.


@node mysql_connect, mysql_create_db, mysql_close, C API functions
@subsubsection @code{mysql_connect()}

@findex @code{mysql_connect()}

@code{MYSQL *mysql_connect(MYSQL *mysql, const char *host, const char *user, const char *passwd)}

@subsubheading Description

This function is deprecated.  It is preferable to use
@code{mysql_real_connect()} instead.

@code{mysql_connect()} attempts to establish a connection to a MySQL
database engine running on @code{host}.  @code{mysql_connect()} must complete
successfully before you can execute any of the other API functions, with the
exception of @code{mysql_get_client_info()}.

The meanings of the parameters are the same as for the corresponding
parameters for @code{mysql_real_connect()} with the difference that the
connection parameter may be @code{NULL}. In this case the C API
allocates memory for the connection structure automatically and frees it
when you call @code{mysql_close()}. The disadvantage of this approach is
that you can't retrieve an error message if the connection fails.  (To
get error information from @code{mysql_errno()} or @code{mysql_error()},
you must provide a valid @code{MYSQL} pointer.)

@subsubheading Return Values

Same as for @code{mysql_real_connect()}.

@subsubheading Errors

Same as for @code{mysql_real_connect()}.


@node mysql_create_db, mysql_data_seek, mysql_connect, C API functions
@subsubsection @code{mysql_create_db()}

@findex @code{mysql_create_db()}

@code{int mysql_create_db(MYSQL *mysql, const char *db)}

@subsubheading Description
Creates the database named by the @code{db} parameter.

This function is deprecated.  It is preferable to use @code{mysql_query()}
to issue a SQL @code{CREATE DATABASE} statement instead.

@subsubheading Return Values

Zero if the database was created successfully. Non-zero if an error
occurred.

@subsubheading Errors
@table @code

@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.

@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.

@item CR_SERVER_LOST
The connection to the server was lost during the query.

@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table

@subsubheading Example

@example
if(mysql_create_db(&mysql, "my_database"))
@{
   fprintf(stderr, "Failed to create new database.  Error: %s\n",
           mysql_error(&mysql));
@}
@end example


@node mysql_data_seek, mysql_debug, mysql_create_db, C API functions
@subsubsection @code{mysql_data_seek()}

@findex @code{mysql_data_seek()}

@code{void mysql_data_seek(MYSQL_RES *result, my_ulonglong offset)}

@subsubheading Description
Seeks to an arbitrary row in a query result set.  This requires that the
result set structure contains the entire result of the query, so
@code{mysql_data_seek()} may be used in conjunction only with
@code{mysql_store_result()}, not with @code{mysql_use_result()}.

The offset should be a value in the range from 0 to
@code{mysql_num_rows(result)-1}.

@subsubheading Return Values

None.

@subsubheading Errors
None.


@node mysql_debug, mysql_drop_db, mysql_data_seek, C API functions
@subsubsection @code{mysql_debug()}

@findex @code{mysql_debug()}

@code{void mysql_debug(const char *debug)}

@subsubheading Description
Does a @code{DBUG_PUSH} with the given string. @code{mysql_debug()} uses the
Fred Fish debug library. To use this function, you must compile the client
library to support debugging.
@xref{Debugging server}. @xref{Debugging client}.

@subsubheading Return Values

None.

@subsubheading Errors
None.

@subsubheading Example

The call shown here causes the client library to generate a trace file in
@file{/tmp/client.trace} on the client machine:

@example
mysql_debug("d:t:O,/tmp/client.trace");
@end example


@node mysql_drop_db, mysql_dump_debug_info, mysql_debug, C API functions
@subsubsection @code{mysql_drop_db()}

@findex @code{mysql_drop_db()}

@code{int mysql_drop_db(MYSQL *mysql, const char *db)}

@subsubheading Description
Drops the database named by the @code{db} parameter.

This function is deprecated.  It is preferable to use @code{mysql_query()}
to issue a SQL @code{DROP DATABASE} statement instead.

@subsubheading Return Values

Zero if the database was dropped successfully. Non-zero if an error
occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table

@subsubheading Example

@example
if(mysql_drop_db(&mysql, "my_database"))
  fprintf(stderr, "Failed to drop the database: Error: %s\n",
          mysql_error(&mysql));
@end example


@node mysql_dump_debug_info, mysql_eof, mysql_drop_db, C API functions
@subsubsection @code{mysql_dump_debug_info()}

@findex @code{mysql_dump_debug_info()}

@code{int mysql_dump_debug_info(MYSQL *mysql)}

@subsubheading Description

Instructs the server to write some debug information to the log.  The
connected user must have the @strong{process} privilege for this to work.

@subsubheading Return Values

Zero if the command was successful. Non-zero if an error occurred.

@subsubheading Errors
@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_eof, mysql_errno, mysql_dump_debug_info, C API functions
@subsubsection @code{mysql_eof()}

@findex @code{mysql_eof()}

@code{my_bool mysql_eof(MYSQL_RES *result)}

@subsubheading Description

This function is deprecated.  @code{mysql_errno()} or @code{mysql_error()}
may be used instead.

@code{mysql_eof()} determines whether the last row of a result
set has been read.

If you acquire a result set from a successful call to
@code{mysql_store_result()}, the client receives the entire set in one
operation.  In this case, a @code{NULL} return from
@code{mysql_fetch_row()} always means the end of the result set has been
reached and it is unnecessary to call @code{mysql_eof()}.  When used
with @code{mysql_store_result()}, @code{mysql_eof()} will always return
true.

On the other hand, if you use @code{mysql_use_result()} to initiate a result
set retrieval, the rows of the set are obtained from the server one by one as
you call @code{mysql_fetch_row()} repeatedly.  Because an error may occur on
the connection during this process, a @code{NULL} return value from
@code{mysql_fetch_row()} does not necessarily mean the end of the result set
was reached normally.  In this case, you can use @code{mysql_eof()} to
determine what happened.  @code{mysql_eof()} returns a non-zero value if the
end of the result set was reached and zero if an error occurred.

Historically, @code{mysql_eof()} predates the standard MySQL error
functions @code{mysql_errno()} and @code{mysql_error()}.  Because those error
functions provide the same information, their use is preferred over
@code{mysql_eof()}, which is now deprecated.  (In fact, they provide more
information, because @code{mysql_eof()} returns only a boolean value whereas
the error functions indicate a reason for the error when one occurs.)

@subsubheading Return Values

Zero if no error occurred. Non-zero if the end of the result set has been
reached.

@subsubheading Errors
None.

@subsubheading Example

The following example shows how you might use @code{mysql_eof()}:

@example
mysql_query(&mysql,"SELECT * FROM some_table");
result = mysql_use_result(&mysql);
while((row = mysql_fetch_row(result)))
@{
    // do something with data
@}
if(!mysql_eof(result))  // mysql_fetch_row() failed due to an error
@{
    fprintf(stderr, "Error: %s\n", mysql_error(&mysql));
@}
@end example

However, you can achieve the same effect with the standard MySQL
error functions:

@example
mysql_query(&mysql,"SELECT * FROM some_table");
result = mysql_use_result(&mysql);
while((row = mysql_fetch_row(result)))
@{
    // do something with data
@}
if(mysql_errno(&mysql))  // mysql_fetch_row() failed due to an error
@{
    fprintf(stderr, "Error: %s\n", mysql_error(&mysql));
@}
@end example


@node mysql_errno, mysql_error, mysql_eof, C API functions
@subsubsection @code{mysql_errno()}

@findex @code{mysql_errno()}

@code{unsigned int mysql_errno(MYSQL *mysql)}

@subsubheading Description

For the connection specified by @code{mysql}, @code{mysql_errno()} returns
the error code for the most recently invoked API function that can succeed
or fail.  A return value of zero means that no error occurred. Client error
message numbers are listed in the MySQL @file{errmsg.h} header file.
Server error message numbers are listed in @file{mysqld_error.h}. In the
MySQL source distribution you can find a complete list of
error messages and error numbers in the file @file{Docs/mysqld_error.txt}.

@subsubheading Return Values

An error code value. Zero if no error occurred.

@subsubheading Errors
None.


@node mysql_error, mysql_escape_string, mysql_errno, C API functions
@subsubsection @code{mysql_error()}

@findex @code{mysql_error()}

@code{char *mysql_error(MYSQL *mysql)}

@subsubheading Description

For the connection specified by @code{mysql}, @code{mysql_error()} returns
the error message for the most recently invoked API function that can succeed
or fail.  An empty string (@code{""}) is returned if no error occurred.
This means the following two tests are equivalent:

@example
if(mysql_errno(&mysql))
@{
    // an error occurred
@}

if(mysql_error(&mysql)[0] != '\0')
@{
    // an error occurred
@}
@end example

The language of the client error messages may be changed by
recompiling the MySQL client library. Currently you can choose
error messages in several different languages.
@xref{Languages}.

@subsubheading Return Values

A character string that describes the error.  An empty string if no error
occurred.

@subsubheading Errors
None.


@node mysql_escape_string, mysql_fetch_field, mysql_error, C API functions
@subsubsection @code{mysql_escape_string()}

@findex @code{mysql_escape_string()}

You should use @code{mysql_real_escape_string()} instead!

This function is identical to @code{mysql_real_escape_string()} except
that @code{mysql_real_escape_string()} takes a connection handler as
its first argument and escapes the string according to the current
character set.  @code{mysql_escape_string()} does not take a connection
argument and does not respect the current charset setting.


@node mysql_fetch_field, mysql_fetch_fields, mysql_escape_string, C API functions
@subsubsection @code{mysql_fetch_field()}

@findex @code{mysql_fetch_field()}

@code{MYSQL_FIELD *mysql_fetch_field(MYSQL_RES *result)}

@subsubheading Description

Returns the definition of one column of a result set as a @code{MYSQL_FIELD}
structure. Call this function repeatedly to retrieve information about all
columns in the result set.  @code{mysql_fetch_field()} returns @code{NULL}
when no more fields are left.

@code{mysql_fetch_field()} is reset to return information about the first
field each time you execute a new @code{SELECT} query.  The field returned by
@code{mysql_fetch_field()} is also affected by calls to
@code{mysql_field_seek()}.

If you've called @code{mysql_query()} to perform a @code{SELECT} on a table
but have not called @code{mysql_store_result()}, MySQL returns the
default blob length (8K bytes) if you call @code{mysql_fetch_field()} to ask
for the length of a @code{BLOB} field.  (The 8K size is chosen because
MySQL doesn't know the maximum length for the @code{BLOB}. This
should be made configurable sometime.) Once you've retrieved the result set,
@code{field->max_length} contains the length of the largest value for this
column in the specific query.

@subsubheading Return Values

The @code{MYSQL_FIELD} structure for the current column. @code{NULL}
if no columns are left.

@subsubheading Errors
None.

@subsubheading Example

@example
MYSQL_FIELD *field;

while((field = mysql_fetch_field(result)))
@{
    printf("field name %s\n", field->name);
@}
@end example


@node mysql_fetch_fields, mysql_fetch_field_direct, mysql_fetch_field, C API functions
@subsubsection @code{mysql_fetch_fields()}

@findex @code{mysql_fetch_fields()}

@code{MYSQL_FIELD *mysql_fetch_fields(MYSQL_RES *result)}

@subsubheading Description

Returns an array of all @code{MYSQL_FIELD} structures for a result set.
Each structure provides the field definition for one column of the result
set.

@subsubheading Return Values

An array of @code{MYSQL_FIELD} structures for all columns of a result set.

@subsubheading Errors
None.

@subsubheading Example

@example
unsigned int num_fields;
unsigned int i;
MYSQL_FIELD *fields;

num_fields = mysql_num_fields(result);
fields = mysql_fetch_fields(result);
for(i = 0; i < num_fields; i++)
@{
   printf("Field %u is %s\n", i, fields[i].name);
@}
@end example


@node mysql_fetch_field_direct, mysql_fetch_lengths, mysql_fetch_fields, C API functions
@subsubsection @code{mysql_fetch_field_direct()}

@findex @code{mysql_fetch_field_direct()}

@code{MYSQL_FIELD *mysql_fetch_field_direct(MYSQL_RES *result, unsigned int fieldnr)}

@subsubheading Description

Given a field number @code{fieldnr} for a column within a result set, returns
that column's field definition as a @code{MYSQL_FIELD} structure. You may use
this function to retrieve the definition for an arbitrary column.  The value
of @code{fieldnr} should be in the range from 0 to
@code{mysql_num_fields(result)-1}.

@subsubheading Return Values

The @code{MYSQL_FIELD} structure for the specified column.

@subsubheading Errors
None.

@subsubheading Example

@example
unsigned int num_fields;
unsigned int i;
MYSQL_FIELD *field;

num_fields = mysql_num_fields(result);
for(i = 0; i < num_fields; i++)
@{
    field = mysql_fetch_field_direct(result, i);
    printf("Field %u is %s\n", i, field->name);
@}
@end example


@node mysql_fetch_lengths, mysql_fetch_row, mysql_fetch_field_direct, C API functions
@subsubsection @code{mysql_fetch_lengths()}

@findex @code{mysql_fetch_lengths()}

@code{unsigned long *mysql_fetch_lengths(MYSQL_RES *result)}

@subsubheading Description

Returns the lengths of the columns of the current row within a result set.
If you plan to copy field values, this length information is also useful for
optimisation, because you can avoid calling @code{strlen()}.  In addition, if
the result set contains binary data, you @strong{must} use this function to
determine the size of the data, because @code{strlen()} returns incorrect
results for any field containing null characters.

The length for empty columns and for columns containing @code{NULL} values is
zero.  To see how to distinguish these two cases, see the description for
@code{mysql_fetch_row()}.

@subsubheading Return Values

An array of unsigned long integers representing the size of each column (not
including any terminating null characters).
@code{NULL} if an error occurred.

@subsubheading Errors
@code{mysql_fetch_lengths()} is valid only for the current row of the result
set.  It returns @code{NULL} if you call it before calling
@code{mysql_fetch_row()} or after retrieving all rows in the result.

@subsubheading Example

@example
MYSQL_ROW row;
unsigned long *lengths;
unsigned int num_fields;
unsigned int i;

row = mysql_fetch_row(result);
if (row)
@{
    num_fields = mysql_num_fields(result);
    lengths = mysql_fetch_lengths(result);
    for(i = 0; i < num_fields; i++)
    @{
         printf("Column %u is %lu bytes in length.\n", i, lengths[i]);
    @}
@}
@end example


@node mysql_fetch_row, mysql_field_count, mysql_fetch_lengths, C API functions
@subsubsection @code{mysql_fetch_row()}

@findex @code{mysql_fetch_row()}

@code{MYSQL_ROW mysql_fetch_row(MYSQL_RES *result)}

@subsubheading Description

Retrieves the next row of a result set. When used after
@code{mysql_store_result()}, @code{mysql_fetch_row()} returns @code{NULL}
when there are no more rows to retrieve.  When used after
@code{mysql_use_result()}, @code{mysql_fetch_row()} returns @code{NULL} when
there are no more rows to retrieve or if an error occurred.

The number of values in the row is given by @code{mysql_num_fields(result)}.
If @code{row} holds the return value from a call to @code{mysql_fetch_row()},
pointers to the values are accessed as @code{row[0]} to
@code{row[mysql_num_fields(result)-1]}.  @code{NULL} values in the row are
indicated by @code{NULL} pointers.

The lengths of the field values in the row may be obtained by calling
@code{mysql_fetch_lengths()}.  Empty fields and fields containing
@code{NULL} both have length 0; you can distinguish these by checking
the pointer for the field value.  If the pointer is @code{NULL}, the field
is @code{NULL}; otherwise, the field is empty.

@subsubheading Return Values

A @code{MYSQL_ROW} structure for the next row.  @code{NULL} if
there are no more rows to retrieve or if an error occurred.

@subsubheading Errors

@table @code
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table

@subsubheading Example

@example
MYSQL_ROW row;
unsigned int num_fields;
unsigned int i;

num_fields = mysql_num_fields(result);
while ((row = mysql_fetch_row(result)))
@{
   unsigned long *lengths;
   lengths = mysql_fetch_lengths(result);
   for(i = 0; i < num_fields; i++)
   @{
       printf("[%.*s] ", (int) lengths[i], row[i] ? row[i] : "NULL");
   @}
   printf("\n");
@}
@end example


@node mysql_field_count, mysql_field_seek, mysql_fetch_row, C API functions
@subsubsection @code{mysql_field_count()}

@findex @code{mysql_field_count()}

@code{unsigned int mysql_field_count(MYSQL *mysql)}

If you are using a version of MySQL earlier than Version 3.22.24, you
should use @code{unsigned int mysql_num_fields(MYSQL *mysql)} instead.

@subsubheading Description

Returns the number of columns for the most recent query on the connection.

The normal use of this function is when @code{mysql_store_result()}
returned @code{NULL} (and thus you have no result set pointer).
In this case, you can call @code{mysql_field_count()} to
determine whether @code{mysql_store_result()} should have produced a
non-empty result.  This allows the client program to take proper action
without knowing whether the query was a @code{SELECT} (or
@code{SELECT}-like) statement.  The example shown here illustrates how this
may be done.

@xref{NULL mysql_store_result, , @code{NULL mysql_store_result()}}.

@subsubheading Return Values

An unsigned integer representing the number of fields in a result set.

@subsubheading Errors
None.

@subsubheading Example

@example
MYSQL_RES *result;
unsigned int num_fields;
unsigned int num_rows;

if (mysql_query(&mysql,query_string))
@{
    // error
@}
else // query succeeded, process any data returned by it
@{
    result = mysql_store_result(&mysql);
    if (result)  // there are rows
    @{
        num_fields = mysql_num_fields(result);
        // retrieve rows, then call mysql_free_result(result)
    @}
    else  // mysql_store_result() returned nothing; should it have?
    @{
        if(mysql_field_count(&mysql) == 0)
        @{
            // query does not return data
            // (it was not a SELECT)
            num_rows = mysql_affected_rows(&mysql);
        @}
        else // mysql_store_result() should have returned data
        @{
            fprintf(stderr, "Error: %s\n", mysql_error(&mysql));
        @}
    @}
@}
@end example

An alternative is to replace the @code{mysql_field_count(&mysql)} call with
@code{mysql_errno(&mysql)}.  In this case, you are checking directly for an
error from @code{mysql_store_result()} rather than inferring from the value
of @code{mysql_field_count()} whether the statement was a
@code{SELECT}.


@node mysql_field_seek, mysql_field_tell, mysql_field_count, C API functions
@subsubsection @code{mysql_field_seek()}

@findex @code{mysql_field_seek()}

@code{MYSQL_FIELD_OFFSET mysql_field_seek(MYSQL_RES *result, MYSQL_FIELD_OFFSET offset)}

@subsubheading Description

Sets the field cursor to the given offset.   The next call to
@code{mysql_fetch_field()} will retrieve the field definition of the column
associated with that offset.

To seek to the beginning of a row, pass an @code{offset} value of zero.

@subsubheading Return Values

The previous value of the field cursor.

@subsubheading Errors
None.


@node mysql_field_tell, mysql_free_result, mysql_field_seek, C API functions
@subsubsection @code{mysql_field_tell()}

@findex @code{mysql_field_tell()}

@code{MYSQL_FIELD_OFFSET mysql_field_tell(MYSQL_RES *result)}

@subsubheading Description

Returns the position of the field cursor used for the last
@code{mysql_fetch_field()}. This value can be used as an argument to
@code{mysql_field_seek()}.

@subsubheading Return Values

The current offset of the field cursor.

@subsubheading Errors
None.


@node mysql_free_result, mysql_get_client_info, mysql_field_tell, C API functions
@subsubsection @code{mysql_free_result()}

@findex @code{mysql_free_result()}

@code{void mysql_free_result(MYSQL_RES *result)}

@subsubheading Description

Frees the memory allocated for a result set by @code{mysql_store_result()},
@code{mysql_use_result()}, @code{mysql_list_dbs()}, etc.  When you are done
with a result set, you must free the memory it uses by calling
@code{mysql_free_result()}.

@subsubheading Return Values

None.

@subsubheading Errors
None.


@node mysql_get_client_info, mysql_get_host_info, mysql_free_result, C API functions
@subsubsection @code{mysql_get_client_info()}

@findex @code{mysql_get_client_info()}

@code{char *mysql_get_client_info(void)}

@subsubheading Description

Returns a string that represents the client library version.

@subsubheading Return Values

A character string that represents the MySQL client library version.

@subsubheading Errors
None.


@node mysql_get_host_info, mysql_get_proto_info, mysql_get_client_info, C API functions
@subsubsection @code{mysql_get_host_info()}

@findex @code{mysql_get_host_info()}

@code{char *mysql_get_host_info(MYSQL *mysql)}

@subsubheading Description

Returns a string describing the type of connection in use, including the
server host name.

@subsubheading Return Values

A character string representing the server host name and the connection type.

@subsubheading Errors
None.


@node mysql_get_proto_info, mysql_get_server_info, mysql_get_host_info, C API functions
@subsubsection @code{mysql_get_proto_info()}

@findex @code{mysql_get_proto_info()}

@code{unsigned int mysql_get_proto_info(MYSQL *mysql)}

@subsubheading Description

Returns the protocol version used by current connection.

@subsubheading Return Values

An unsigned integer representing the protocol version used by the current
connection.

@subsubheading Errors
None.


@node mysql_get_server_info, mysql_info, mysql_get_proto_info, C API functions
@subsubsection @code{mysql_get_server_info()}

@findex @code{mysql_get_server_info()}

@code{char *mysql_get_server_info(MYSQL *mysql)}

@subsubheading Description

Returns a string that represents the server version number.

@subsubheading Return Values

A character string that represents the server version number.

@subsubheading Errors
None.


@node mysql_info, mysql_init, mysql_get_server_info, C API functions
@subsubsection @code{mysql_info()}

@findex @code{mysql_info()}

@code{char *mysql_info(MYSQL *mysql)}

@subsubheading Description

Retrieves a string providing information about the most recently executed
query, but only for the statements listed here.  For other statements,
@code{mysql_info()} returns @code{NULL}.  The format of the string varies
depending on the type of query, as described here.  The numbers are
illustrative only; the string will contain values appropriate for the query.

@table @code
@item INSERT INTO ... SELECT ...
String format: @code{Records: 100 Duplicates: 0 Warnings: 0}
@item INSERT INTO ... VALUES (...),(...),(...)...
String format: @code{Records: 3  Duplicates: 0  Warnings: 0}
@item LOAD DATA INFILE ...
String format: @code{Records: 1 Deleted: 0 Skipped: 0 Warnings: 0}
@item ALTER TABLE
String format: @code{Records: 3 Duplicates: 0 Warnings: 0}
@item UPDATE
String format: @code{Rows matched: 40 Changed: 40 Warnings: 0}
@end table

Note that @code{mysql_info()} returns a non-@code{NULL} value for the
@code{INSERT ... VALUES} statement only if multiple value lists are
specified in the statement.

@subsubheading Return Values

A character string representing additional information about the most
recently executed query.  @code{NULL} if no information is available for the
query.

@subsubheading Errors
None.


@node mysql_init, mysql_insert_id, mysql_info, C API functions
@subsubsection @code{mysql_init()}

@findex @code{mysql_init()}

@code{MYSQL *mysql_init(MYSQL *mysql)}

@subsubheading Description

Allocates or initialises a @code{MYSQL} object suitable for
@code{mysql_real_connect()}.  If @code{mysql} is a @code{NULL} pointer, the
function allocates, initialises, and returns a new object.  Otherwise, the
object is initialised and the address of the object is returned.  If
@code{mysql_init()} allocates a new object, it will be freed when
@code{mysql_close()} is called to close the connection.

@subsubheading Return Values

An initialised @code{MYSQL*} handle.  @code{NULL} if there was
insufficient memory to allocate a new object.

@subsubheading Errors
In case of insufficient memory, @code{NULL} is returned.


@node mysql_insert_id, mysql_kill, mysql_init, C API functions
@subsubsection @code{mysql_insert_id()}

@findex @code{mysql_insert_id()}

@code{my_ulonglong mysql_insert_id(MYSQL *mysql)}

@subsubheading Description

Returns the ID generated for an @code{AUTO_INCREMENT} column by the previous
query.  Use this function after you have performed an @code{INSERT} query
into a table that contains an @code{AUTO_INCREMENT} field.

Note that @code{mysql_insert_id()} returns @code{0} if the previous query
does not generate an @code{AUTO_INCREMENT} value.  If you need to save
the value for later, be sure to call @code{mysql_insert_id()} immediately
after the query that generates the value.

@code{mysql_insert_id()} is updated after @code{INSERT} and
@code{UPDATE} statements that generate an @code{AUTO_INCREMENT} value or
that set a column value to @code{LAST_INSERT_ID(expr)}.
@xref{Miscellaneous functions}.

Also note that the value of the SQL @code{LAST_INSERT_ID()} function always
contains the most recently generated @code{AUTO_INCREMENT} value, and is
not reset between queries because the value of that function is maintained
in the server.

@subsubheading Return Values

The value of the @code{AUTO_INCREMENT} field that was updated by the previous
query.  Returns zero if there was no previous query on the connection or if
the query did not update an @code{AUTO_INCREMENT} value.

@subsubheading Errors
None.


@node mysql_kill, mysql_list_dbs, mysql_insert_id, C API functions
@subsubsection @code{mysql_kill()}

@findex @code{mysql_kill()}

@code{int mysql_kill(MYSQL *mysql, unsigned long pid)}

@subsubheading Description

Asks the server to kill the thread specified by @code{pid}.

@subsubheading Return Values

Zero for success.  Non-zero if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_list_dbs, mysql_list_fields, mysql_kill, C API functions
@subsubsection @code{mysql_list_dbs()}

@findex @code{mysql_list_dbs()}

@code{MYSQL_RES *mysql_list_dbs(MYSQL *mysql, const char *wild)}

@subsubheading Description

Returns a result set consisting of database names on the server that match
the simple regular expression specified by the @code{wild} parameter.
@code{wild} may contain the wildcard characters @samp{%} or @samp{_}, or may
be a @code{NULL} pointer to match all databases.  Calling
@code{mysql_list_dbs()} is similar to executing the query @code{SHOW
databases [LIKE wild]}.

You must free the result set with @code{mysql_free_result()}.

@subsubheading Return Values

A @code{MYSQL_RES} result set for success.  @code{NULL} if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_OUT_OF_MEMORY
Out of memory.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_list_fields, mysql_list_processes, mysql_list_dbs, C API functions
@subsubsection  @code{mysql_list_fields()}

@findex @code{mysql_list_fields()}

@code{MYSQL_RES *mysql_list_fields(MYSQL *mysql, const char *table, const char *wild)}

@subsubheading Description

Returns a result set consisting of field names in the given table that match
the simple regular expression specified by the @code{wild} parameter.
@code{wild} may contain the wildcard characters @samp{%} or @samp{_}, or may
be a @code{NULL} pointer to match all fields.  Calling
@code{mysql_list_fields()} is similar to executing the query @code{SHOW
COLUMNS FROM tbl_name [LIKE wild]}.

Note that it's recommended that you use @code{SHOW COLUMNS FROM tbl_name}
instead of @code{mysql_list_fields()}.

You must free the result set with @code{mysql_free_result()}.

@subsubheading Return Values

A @code{MYSQL_RES} result set for success.  @code{NULL} if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_list_processes, mysql_list_tables, mysql_list_fields, C API functions
@subsubsection @code{mysql_list_processes()}

@findex @code{mysql_list_processes()}

@code{MYSQL_RES *mysql_list_processes(MYSQL *mysql)}

@subsubheading Description

Returns a result set describing the current server threads.  This is the same
kind of information as that reported by @code{mysqladmin processlist} or
a @code{SHOW PROCESSLIST} query.

You must free the result set with @code{mysql_free_result()}.

@subsubheading Return Values

A @code{MYSQL_RES} result set for success.  @code{NULL} if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_list_tables, mysql_num_fields, mysql_list_processes, C API functions
@subsubsection @code{mysql_list_tables()}

@findex @code{mysql_list_tables()}

@code{MYSQL_RES *mysql_list_tables(MYSQL *mysql, const char *wild)}

@subsubheading Description

Returns a result set consisting of table names in the current database that
match the simple regular expression specified by the @code{wild} parameter.
@code{wild} may contain the wildcard characters @samp{%} or @samp{_}, or may
be a @code{NULL} pointer to match all tables.  Calling
@code{mysql_list_tables()} is similar to executing the query @code{SHOW
tables [LIKE wild]}.

You must free the result set with @code{mysql_free_result()}.

@subsubheading Return Values

A @code{MYSQL_RES} result set for success.  @code{NULL} if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_num_fields, mysql_num_rows, mysql_list_tables, C API functions
@subsubsection @code{mysql_num_fields()}

@findex @code{mysql_num_fields()}
@findex @code{mysql_field_count()}

@code{unsigned int mysql_num_fields(MYSQL_RES *result)}

or

@code{unsigned int mysql_num_fields(MYSQL *mysql)}

The second form doesn't work on MySQL Version 3.22.24 or newer.  To pass a
@code{MYSQL*} argument, you must use
@code{unsigned int mysql_field_count(MYSQL *mysql)} instead.

@subsubheading Description

Returns the number of columns in a result set.

Note that you can get the number of columns either from a pointer to a result
set or to a connection handle.  You would use the connection handle if
@code{mysql_store_result()} or @code{mysql_use_result()} returned
@code{NULL} (and thus you have no result set pointer).  In this case, you can
call @code{mysql_field_count()} to determine whether
@code{mysql_store_result()} should have produced a non-empty result.  This
allows the client program to take proper action without knowing whether or
not the query was a @code{SELECT} (or @code{SELECT}-like) statement.  The
example shown here illustrates how this may be done.

@xref{NULL mysql_store_result, , @code{NULL mysql_store_result()}}.

@subsubheading Return Values

An unsigned integer representing the number of fields in a result set.

@subsubheading Errors
None.

@subsubheading Example

@example
MYSQL_RES *result;
unsigned int num_fields;
unsigned int num_rows;

if (mysql_query(&mysql,query_string))
@{
    // error
@}
else // query succeeded, process any data returned by it
@{
    result = mysql_store_result(&mysql);
    if (result)  // there are rows
    @{
        num_fields = mysql_num_fields(result);
        // retrieve rows, then call mysql_free_result(result)
    @}
    else  // mysql_store_result() returned nothing; should it have?
    @{
        if (mysql_errno(&mysql))
	@{
           fprintf(stderr, "Error: %s\n", mysql_error(&mysql));
	@}
        else if (mysql_field_count(&mysql) == 0)
        @{
            // query does not return data
            // (it was not a SELECT)
            num_rows = mysql_affected_rows(&mysql);
        @}
    @}
@}
@end example

An alternative (if you know that your query should have returned a result set)
is to replace the @code{mysql_errno(&mysql)} call with a check if
@code{mysql_field_count(&mysql)} is = 0.  This will only happen if something
went wrong.


@node mysql_num_rows, mysql_options, mysql_num_fields, C API functions
@subsubsection @code{mysql_num_rows()}

@findex @code{mysql_num_rows()}

@code{my_ulonglong mysql_num_rows(MYSQL_RES *result)}

@subsubheading Description

Returns the number of rows in the result set.

The use of @code{mysql_num_rows()} depends on whether you use
@code{mysql_store_result()} or @code{mysql_use_result()} to return the result
set.  If you use @code{mysql_store_result()}, @code{mysql_num_rows()} may be
called immediately.  If you use @code{mysql_use_result()},
@code{mysql_num_rows()} will not return the correct value until all the rows
in the result set have been retrieved.

@subsubheading Return Values

The number of rows in the result set.

@subsubheading Errors
None.


@node mysql_options, mysql_ping, mysql_num_rows, C API functions
@subsubsection @code{mysql_options()}

@findex @code{mysql_options()}

@code{int mysql_options(MYSQL *mysql, enum mysql_option option, const char *arg)}

@subsubheading Description

Can be used to set extra connect options and affect behavior for a connection.
This function may be called multiple times to set several options.

@code{mysql_options()} should be called after @code{mysql_init()} and before
@code{mysql_connect()} or @code{mysql_real_connect()}.

The @code{option} argument is the option that you want to set; the @code{arg}
argument is the value for the option. If the option is an integer, then
@code{arg} should point to the value of the integer.

Possible options values:

@multitable @columnfractions .42 .20 .38
@item @strong{Option} @tab @strong{Argument type} @tab @strong{Function}
@item @code{MYSQL_OPT_CONNECT_TIMEOUT} @tab @code{unsigned int *} @tab Connect timeout in seconds.
@item @code{MYSQL_OPT_COMPRESS} @tab Not used @tab Use the compressed client/server protocol.
@item @code{MYSQL_OPT_LOCAL_INFILE} @tab optional pointer to uint @tab If no pointer is given or if pointer points to an @code{unsigned int != 0} the command @code{LOAD LOCAL INFILE} is enabled.
@item @code{MYSQL_OPT_NAMED_PIPE} @tab Not used @tab Use named pipes to connect to a MySQL server on NT.
@item @code{MYSQL_INIT_COMMAND} @tab @code{char *} @tab Command to execute when connecting to the MySQL server. Will automatically be re-executed when reconnecting.
@item @code{MYSQL_READ_DEFAULT_FILE} @tab @code{char *} @tab Read options from the named option file instead of from @file{my.cnf}.
@item @code{MYSQL_READ_DEFAULT_GROUP} @tab @code{char *} @tab Read options from the named group from @file{my.cnf} or the file specified with @code{MYSQL_READ_DEFAULT_FILE}.
@end multitable

Note that the group @code{client} is always read if you use
@code{MYSQL_READ_DEFAULT_FILE} or @code{MYSQL_READ_DEFAULT_GROUP}.

The specified group in the option file may contain the following options:

@multitable @columnfractions .25 .60
@item @strong{Option} @tab @strong{Description}
@item @code{connect-timeout} @tab Connect timeout in seconds. On Linux this timeout is also used for waiting for the first answer from the server.
@item @code{compress} @tab Use the compressed client/server protocol.
@item @code{database} @tab Connect to this database if no database was specified in the connect command.
@item @code{debug} @tab Debug options.
@item @code{disable-local-infile} @tab Disable use of @code{LOAD DATA LOCAL}.
@item @code{host} @tab Default host name.
@item @code{init-command} @tab Command to execute when connecting to MySQL server. Will automatically be re-executed when reconnecting.
@item @code{interactive-timeout} @tab Same as specifying @code{CLIENT_INTERACTIVE} to @code{mysql_real_connect()}. @xref{mysql_real_connect}.
@item @code{local-infile[=(0|1)]} @tab If no argument or argument != 0 then enable use of @code{LOAD DATA LOCAL}.
@item @code{password} @tab Default password.
@item @code{pipe} @tab Use named pipes to connect to a MySQL server on NT.
@item @code{port} @tab Default port number.
@item @code{return-found-rows} @tab Tell @code{mysql_info()} to return found rows instead of updated rows when using @code{UPDATE}.
@item @code{socket} @tab Default socket number.
@item @code{user} @tab Default user.
@end multitable

Note that @code{timeout} has been replaced by @code{connect-timeout}, but
@code{timeout} will still work for a while.

For more information about option files, see @ref{Option files}.

@subsubheading Return Values

Zero for success.  Non-zero if you used an unknown option.

@subsubheading Example

@example
MYSQL mysql;

mysql_init(&mysql);
mysql_options(&mysql,MYSQL_OPT_COMPRESS,0);
mysql_options(&mysql,MYSQL_READ_DEFAULT_GROUP,"odbc");
if (!mysql_real_connect(&mysql,"host","user","passwd","database",0,NULL,0))
@{
    fprintf(stderr, "Failed to connect to database: Error: %s\n",
          mysql_error(&mysql));
@}
@end example

The above requests the client to use the compressed client/server protocol and
read the additional options from the @code{odbc} section in the @file{my.cnf}
file.


@node mysql_ping, mysql_query, mysql_options, C API functions
@subsubsection @code{mysql_ping()}

@findex @code{mysql_ping()}

@code{int mysql_ping(MYSQL *mysql)}

@subsubheading Description

Checks whether the connection to the server is working. If it has gone
down, an automatic reconnection is attempted.

This function can be used by clients that remain idle for a long while,
to check whether the server has closed the connection and reconnect
if necessary.

@subsubheading Return Values

Zero if the server is alive. Non-zero if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_query, mysql_real_connect, mysql_ping, C API functions
@subsubsection @code{mysql_query()}

@findex @code{mysql_query()}

@code{int mysql_query(MYSQL *mysql, const char *query)}

@subsubheading Description
Executes the SQL query pointed to by the null-terminated string @code{query}.
The query must consist of a single SQL statement.  You should not add
a terminating semicolon (@samp{;}) or @code{\g} to the statement.

@code{mysql_query()} cannot be used for queries that contain binary data; you
should use @code{mysql_real_query()} instead.  (Binary data may contain the
@samp{\0} character, which @code{mysql_query()} interprets as the end of the
query string.)

If you want to know if the query should return a result set or not, you can
use @code{mysql_field_count()} to check for this.
@xref{mysql_field_count, , @code{mysql_field_count()}}.

@subsubheading Return Values

Zero if the query was successful. Non-zero if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_real_connect, mysql_real_escape_string, mysql_query, C API functions
@subsubsection @code{mysql_real_connect()}

@findex @code{mysql_real_connect()}

@code{MYSQL *mysql_real_connect(MYSQL *mysql, const char *host,
                 const char *user, const char *passwd, const char *db,
                 unsigned int port, const char *unix_socket,
                 unsigned int client_flag)}

@subsubheading Description

@code{mysql_real_connect()} attempts to establish a connection to a
MySQL database engine running on @code{host}.
@code{mysql_real_connect()} must complete successfully before you can execute
any of the other API functions, with the exception of
@code{mysql_get_client_info()}.

The parameters are specified as follows:

@itemize @bullet
@item
The first parameter should be the address of an existing @code{MYSQL}
structure.  Before calling @code{mysql_real_connect()} you must call
@code{mysql_init()} to initialise the @code{MYSQL} structure. You can
change a lot of connect options with the @code{mysql_options()}
call.  @xref{mysql_options, @code{mysql_options()}}.

@item
The value of @code{host} may be either a hostname or an IP address.  If
@code{host} is @code{NULL} or the string @code{"localhost"}, a connection to
the local host is assumed. If the OS supports sockets (Unix) or named pipes
(Windows), they are used instead of TCP/IP to connect to the server.

@item
The @code{user} parameter contains the user's MySQL login ID.  If
@code{user} is @code{NULL}, the current user is assumed.  Under Unix, this is
the current login name.  Under Windows ODBC, the current user name must be
specified explicitly.
@xref{ODBC administrator}.

@item
The @code{passwd} parameter contains the password for @code{user}.  If
@code{passwd} is @code{NULL}, only entries in the @code{user} table for the
user that have a blank (empty) password field will be checked for a match. This
allows the database administrator to set up the MySQL privilege
system in such a way that users get different privileges depending on whether
or not they have specified a password.

Note: Do not attempt to encrypt the password before calling
@code{mysql_real_connect()}; password encryption is handled automatically by
the client API.

@item
@code{db} is the database name.
If @code{db} is not @code{NULL}, the connection will set the default
database to this value.

@item
If @code{port} is not 0, the value will be used as the port number
for the TCP/IP connection.  Note that the @code{host} parameter
determines the type of the connection.

@item
If @code{unix_socket} is not @code{NULL}, the string specifies the
socket or named pipe that should be used.  Note that the @code{host}
parameter determines the type of the connection.

@item
The value of @code{client_flag} is usually 0, but can be set to a combination
of the following flags in very special circumstances:

@multitable @columnfractions .30 .65
@item @strong{Flag name}       @tab @strong{Flag description}
@item @code{CLIENT_COMPRESS}   @tab Use compression protocol.
@item @code{CLIENT_FOUND_ROWS} @tab Return the number of found (matched) rows, not the number of affected rows.
@item @code{CLIENT_IGNORE_SPACE} @tab Allow spaces after function names. Makes all functions names reserved words.
@item @code{CLIENT_INTERACTIVE} @tab Allow @code{interactive_timeout} seconds (instead of @code{wait_timeout} seconds) of inactivity before closing the connection.
@item @code{CLIENT_NO_SCHEMA}  @tab Don't allow the @code{db_name.tbl_name.col_name} syntax.  This is for ODBC. It causes the parser to generate an error if you use that syntax, which is useful for trapping bugs in some ODBC programs.
@item @code{CLIENT_ODBC}       @tab The client is an ODBC client. This changes @code{mysqld} to be more ODBC-friendly.
@item @code{CLIENT_SSL}		@tab Use SSL (encrypted protocol).
@end multitable
@end itemize

@subsubheading Return Values

A @code{MYSQL*} connection handle if the connection was successful,
@code{NULL} if the connection was unsuccessful.  For a successful connection,
the return value is the same as the value of the first parameter.

@subsubheading Errors

@table @code
@item CR_CONN_HOST_ERROR
Failed to connect to the MySQL server.

@item CR_CONNECTION_ERROR
Failed to connect to the local MySQL server.

@item CR_IPSOCK_ERROR
Failed to create an IP socket.

@item CR_OUT_OF_MEMORY
Out of memory.

@item CR_SOCKET_CREATE_ERROR
Failed to create a Unix socket.

@item CR_UNKNOWN_HOST
Failed to find the IP address for the hostname.

@item CR_VERSION_ERROR
A protocol mismatch resulted from attempting to connect to a server with a
client library that uses a different protocol version.  This can happen if you
use a very old client library to connect to a new server that wasn't started
with the @code{--old-protocol} option.

@item CR_NAMEDPIPEOPEN_ERROR
Failed to create a named pipe on Windows.

@item CR_NAMEDPIPEWAIT_ERROR
Failed to wait for a named pipe on Windows.

@item CR_NAMEDPIPESETSTATE_ERROR
Failed to get a pipe handler on Windows.

@item CR_SERVER_LOST
If @code{connect_timeout} > 0 and it took longer then @code{connect_timeout}
seconds to connect to the server or if the server died while executing the
@code{init-command}.

@end table

@subsubheading Example

@example
MYSQL mysql;

mysql_init(&mysql);
mysql_options(&mysql,MYSQL_READ_DEFAULT_GROUP,"your_prog_name");
if (!mysql_real_connect(&mysql,"host","user","passwd","database",0,NULL,0))
@{
    fprintf(stderr, "Failed to connect to database: Error: %s\n",
          mysql_error(&mysql));
@}
@end example

By using @code{mysql_options()} the MySQL library will read the
@code{[client]} and @code{your_prog_name} sections in the @file{my.cnf}
file which will ensure that your program will work, even if someone has
set up MySQL in some non-standard way.

Note that upon connection, @code{mysql_real_connect()} sets the @code{reconnect}
flag (part of the MYSQL structure) to a value of @code{1}. This flag indicates,
in the event that a query cannot be performed because of a lost connection, to
try reconnecting to the server before giving up.


@node mysql_real_escape_string, mysql_real_query, mysql_real_connect, C API functions
@subsubsection @code{mysql_real_escape_string()}

@findex @code{mysql_real_escape_string()}

@code{unsigned long mysql_real_escape_string(MYSQL *mysql, char *to, const char *from, unsigned long length)}

@subsubheading Description

This function is used to create a legal SQL string that you can use in a
SQL statement. @xref{String syntax}.

The string in @code{from} is encoded to an escaped SQL string, taking
into account the current character set of the connection. The result is placed
in @code{to} and a terminating null byte is appended.  Characters
encoded are @code{NUL} (ASCII 0), @samp{\n}, @samp{\r}, @samp{\},
@samp{'}, @samp{"}, and Control-Z (@pxref{Literals}).

The string pointed to by @code{from} must be @code{length} bytes long.  You
must allocate the @code{to} buffer to be at least @code{length*2+1} bytes
long.  (In the worse case, each character may need to be encoded as using two
bytes, and you need room for the terminating null byte.)  When
@code{mysql_escape_string()} returns, the contents of @code{to} will be a
null-terminated string.  The return value is the length of the encoded
string, not including the terminating null character.

@subsubheading Example

@example
char query[1000],*end;

end = strmov(query,"INSERT INTO test_table values(");
*end++ = '\'';
end += mysql_real_escape_string(&mysql, end,"What's this",11);
*end++ = '\'';
*end++ = ',';
*end++ = '\'';
end += mysql_real_escape_string(&mysql, end,"binary data: \0\r\n",16);
*end++ = '\'';
*end++ = ')';

if (mysql_real_query(&mysql,query,(unsigned int) (end - query)))
@{
   fprintf(stderr, "Failed to insert row, Error: %s\n",
           mysql_error(&mysql));
@}
@end example

The @code{strmov()} function used in the example is included in the
@code{mysqlclient} library and works like @code{strcpy()} but returns a
pointer to the terminating null of the first parameter.

@subsubheading Return Values

The length of the value placed into @code{to}, not including the
terminating null character.

@subsubheading Errors
None.


@node mysql_real_query, mysql_reload, mysql_real_escape_string, C API functions
@subsubsection @code{mysql_real_query()}

@findex @code{mysql_real_query()}

@code{int mysql_real_query(MYSQL *mysql, const char *query, unsigned long length)}

@subsubheading Description

Executes the SQL query pointed to by @code{query}, which should be a string
@code{length} bytes long.  The query must consist of a single SQL statement.
You should not add a terminating semicolon (@samp{;}) or @code{\g} to the
statement.

You @strong{must} use @code{mysql_real_query()} rather than
@code{mysql_query()} for queries that contain binary data, because binary data
may contain the @samp{\0} character.  In addition, @code{mysql_real_query()}
is faster than @code{mysql_query()} because it does not call @code{strlen()} on
the query string.

If you want to know if the query should return a result set or not, you can
use @code{mysql_field_count()} to check for this.
@xref{mysql_field_count, @code{mysql_field_count()}}.

@subsubheading Return Values

Zero if the query was successful. Non-zero if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_reload, mysql_row_seek, mysql_real_query, C API functions
@subsubsection @code{mysql_reload()}

@findex @code{mysql_reload()}

@code{int mysql_reload(MYSQL *mysql)}

@subsubheading Description

Asks the MySQL server to reload the grant tables. The
connected user must have the @strong{reload} privilege.

This function is deprecated.  It is preferable to use @code{mysql_query()}
to issue a SQL @code{FLUSH PRIVILEGES} statement instead.

@subsubheading Return Values

Zero for success.  Non-zero if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_row_seek, mysql_row_tell, mysql_reload, C API functions
@subsubsection @code{mysql_row_seek()}

@findex @code{mysql_row_seek()}

@code{MYSQL_ROW_OFFSET mysql_row_seek(MYSQL_RES *result, MYSQL_ROW_OFFSET offset)}

@subsubheading Description
Sets the row cursor to an arbitrary row in a query result set.  This requires
that the result set structure contains the entire result of the query, so
@code{mysql_row_seek()} may be used in conjunction only with
@code{mysql_store_result()}, not with @code{mysql_use_result()}.

The offset should be a value returned from a call to @code{mysql_row_tell()}
or to @code{mysql_row_seek()}.  This value is not simply a row number; if you
want to seek to a row within a result set using a row number, use
@code{mysql_data_seek()} instead.

@subsubheading Return Values

The previous value of the row cursor.  This value may be passed to a
subsequent call to @code{mysql_row_seek()}.

@subsubheading Errors
None.


@node mysql_row_tell, mysql_select_db, mysql_row_seek, C API functions
@subsubsection @code{mysql_row_tell()}

@findex @code{mysql_row_tell()}

@code{MYSQL_ROW_OFFSET mysql_row_tell(MYSQL_RES *result)}

@subsubheading Description

Returns the current position of the row cursor for the last
@code{mysql_fetch_row()}. This value can be used as an argument to
@code{mysql_row_seek()}.

You should use @code{mysql_row_tell()} only after @code{mysql_store_result()},
not after @code{mysql_use_result()}.

@subsubheading Return Values

The current offset of the row cursor.

@subsubheading Errors
None.


@node mysql_select_db, mysql_shutdown, mysql_row_tell, C API functions
@subsubsection @code{mysql_select_db()}

@findex @code{mysql_select_db()}

@code{int mysql_select_db(MYSQL *mysql, const char *db)}

@subsubheading Description

Causes the database specified by @code{db} to become the default (current)
database on the connection specified by @code{mysql}.  In subsequent queries,
this database is the default for table references that do not include an
explicit database specifier.

@code{mysql_select_db()} fails unless the connected user can be authenticated
as having permission to use the database.

@subsubheading Return Values

Zero for success.  Non-zero if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_shutdown, mysql_stat, mysql_select_db, C API functions
@subsubsection @code{mysql_shutdown()}

@findex @code{mysql_shutdown()}

@code{int mysql_shutdown(MYSQL *mysql)}

@subsubheading Description

Asks the database server to shut down. The connected user must have
@strong{shutdown} privileges.

@subsubheading Return Values

Zero for success.  Non-zero if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_stat, mysql_store_result, mysql_shutdown, C API functions
@subsubsection @code{mysql_stat()}

@findex @code{mysql_stat()}

@code{char *mysql_stat(MYSQL *mysql)}

@subsubheading Description

Returns a character string containing information similar to that provided by
the @code{mysqladmin status} command.  This includes uptime in seconds and
the number of running threads, questions, reloads, and open tables.

@subsubheading Return Values

A character string describing the server status. @code{NULL} if an
error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_store_result, mysql_thread_id, mysql_stat, C API functions
@subsubsection @code{mysql_store_result()}

@findex @code{mysql_store_result()}

@code{MYSQL_RES *mysql_store_result(MYSQL *mysql)}

@subsubheading Description

You must call @code{mysql_store_result()} or @code{mysql_use_result()}
for every query that successfully retrieves data (@code{SELECT},
@code{SHOW}, @code{DESCRIBE}, @code{EXPLAIN}).

You don't have to call @code{mysql_store_result()} or
@code{mysql_use_result()} for other queries, but it will not do any
harm or cause any notable performance if you call @code{mysql_store_result()}
in all cases.  You can detect if the query didn't have a result set by
checking if @code{mysql_store_result()} returns 0 (more about this later one).

If you want to know if the query should return a result set or not, you can
use @code{mysql_field_count()} to check for this.
@xref{mysql_field_count, @code{mysql_field_count()}}.

@code{mysql_store_result()} reads the entire result of a query to the client,
allocates a @code{MYSQL_RES} structure, and places the result into this
structure.

@code{mysql_store_result()} returns a null pointer if the query didn't return
a result set (if the query was, for example, an @code{INSERT} statement).

@code{mysql_store_result()} also returns a null pointer if reading of the
result set failed. You can check if you got an error by checking if
@code{mysql_error()} doesn't return a null pointer, if
@code{mysql_errno()} returns <> 0, or if @code{mysql_field_count()}
returns <> 0.

An empty result set is returned if there are no rows returned.  (An empty
result set differs from a null pointer as a return value.)

Once you have called @code{mysql_store_result()} and got a result back
that isn't a null pointer, you may call @code{mysql_num_rows()} to find
out how many rows are in the result set.

You can call @code{mysql_fetch_row()} to fetch rows from the result set,
or @code{mysql_row_seek()} and @code{mysql_row_tell()} to obtain or
set the current row position within the result set.

You must call @code{mysql_free_result()} once you are done with the result
set.

@xref{NULL mysql_store_result, , @code{NULL mysql_store_result()}}.

@subsubheading Return Values

A @code{MYSQL_RES} result structure with the results. @code{NULL} if
an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_OUT_OF_MEMORY
Out of memory.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table


@node mysql_thread_id, mysql_use_result, mysql_store_result, C API functions
@subsubsection @code{mysql_thread_id()}

@findex @code{mysql_thread_id()}

@code{unsigned long mysql_thread_id(MYSQL *mysql)}

@subsubheading Description

Returns the thread ID of the current connection.  This value can be used as
an argument to @code{mysql_kill()} to kill the thread.

If the connection is lost and you reconnect with @code{mysql_ping()}, the
thread ID will change.  This means you should not get the thread ID and store
it for later. You should get it when you need it.

@subsubheading Return Values

The thread ID of the current connection.

@subsubheading Errors
None.


@node mysql_use_result,  , mysql_thread_id, C API functions
@subsubsection @code{mysql_use_result()}

@findex @code{mysql_use_result()}

@code{MYSQL_RES *mysql_use_result(MYSQL *mysql)}

@subsubheading Description

You must call @code{mysql_store_result()} or @code{mysql_use_result()} for
every query that successfully retrieves data (@code{SELECT}, @code{SHOW},
@code{DESCRIBE}, @code{EXPLAIN}).

@code{mysql_use_result()} initiates a result set retrieval but does not
actually read the result set into the client like @code{mysql_store_result()}
does.  Instead, each row must be retrieved individually by making calls to
@code{mysql_fetch_row()}.  This reads the result of a query directly from the
server without storing it in a temporary table or local buffer, which is
somewhat faster and uses much less memory than @code{mysql_store_result()}.
The client will only allocate memory for the current row and a communication
buffer that may grow up to @code{max_allowed_packet} bytes.

On the other hand, you shouldn't use @code{mysql_use_result()} if you are
doing a lot of processing for each row on the client side, or if the output
is sent to a screen on which the user may type a @code{^S} (stop scroll).
This will tie up the server and prevent other threads from updating any
tables from which the data is being fetched.

When using @code{mysql_use_result()}, you must execute
@code{mysql_fetch_row()} until a @code{NULL} value is returned, otherwise, the
unfetched rows will be returned as part of the result set for your next
query. The C API will give the error @code{Commands out of sync; you can't
run this command now} if you forget to do this!

You may not use @code{mysql_data_seek()}, @code{mysql_row_seek()},
@code{mysql_row_tell()}, @code{mysql_num_rows()}, or
@code{mysql_affected_rows()} with a result returned from
@code{mysql_use_result()}, nor may you issue other queries until the
@code{mysql_use_result()} has finished.  (However, after you have fetched all
the rows, @code{mysql_num_rows()} will accurately return the number of rows
fetched.)

You must call @code{mysql_free_result()} once you are done with the result
set.

@subsubheading Return Values

A @code{MYSQL_RES} result structure. @code{NULL} if an error occurred.

@subsubheading Errors

@table @code
@item CR_COMMANDS_OUT_OF_SYNC
Commands were executed in an improper order.
@item CR_OUT_OF_MEMORY
Out of memory.
@item CR_SERVER_GONE_ERROR
The MySQL server has gone away.
@item CR_SERVER_LOST
The connection to the server was lost during the query.
@item CR_UNKNOWN_ERROR
An unknown error occurred.
@end table

@node C Thread functions, C Embedded Server func, C API functions, C
@subsection C Threaded Function Descriptions

You need to use the following functions when you want to create a
threaded client. @xref{Threaded clients}.

@menu
* my_init::                     @code{my_init()}
* mysql_thread_init::           @code{mysql_thread_init()}
* mysql_thread_end::            @code{mysql_thread_end()}
* mysql_thread_safe::           @code{mysql_thread_safe()}
@end menu

@node my_init, mysql_thread_init, C Thread functions, C Thread functions
@subsubsection @code{my_init()}

@findex @code{my_init()}

@code{void my_init(void)}

@subsubheading Description

This function needs to be called once in the program before calling any
MySQL function.  This initialises some global variables that MySQL
needs.  If you are using a thread-safe client library, this will also
call @code{mysql_thread_init()} for this thread.

This is automatically called by @code{mysql_init()},
@code{mysql_server_init()} and @code{mysql_connect()}.

@subsubheading Return Values

None.

@node mysql_thread_init, mysql_thread_end, my_init, C Thread functions
@subsubsection @code{mysql_thread_init()}

@findex @code{mysql_thread_init()}

@code{my_bool mysql_thread_init(void)}

@subsubheading Description

This function needs to be called for each created thread to initialise
thread specific variables.

This is automatically called by @code{my_init()} and @code{mysql_connect()}.

@subsubheading Return Values

None.

@node mysql_thread_end, mysql_thread_safe, mysql_thread_init, C Thread functions
@subsubsection @code{mysql_thread_end()}

@findex @code{mysql_thread_end()}

@code{void mysql_thread_end(void)}

@subsubheading Description

This function needs to be called before calling @code{pthread_exit()} to
free memory allocated by @code{mysql_thread_init()}.

Note that this function @strong{is not invoked automatically} by the client
library.  It must be called explicitly to avoid a memory leak.

@subsubheading Return Values

None.

@node mysql_thread_safe,  , mysql_thread_end, C Thread functions
@subsubsection @code{mysql_thread_safe()}

@findex @code{mysql_thread_safe()}

@code{unsigned int mysql_thread_safe(void)}

@subsubheading Description

This function indicates whether the client is compiled as thread-safe.

@subsubheading Return Values

1 is the client is thread-safe, 0 otherwise.

@node C Embedded Server func, C API problems, C Thread functions, C
@subsection C Embedded Server Function Descriptions

You must use the following functions if you want to allow your
application to be linked against the embedded MySQL server library.
@xref{libmysqld}.

If the program is linked with @code{-lmysqlclient} instead of
@code{-lmysqld}, these functions do nothing.  This makes it
possible to choose between using the embedded MySQL server and
a stand-alone server without modifying any code.

@menu
* mysql_server_init::           @code{mysql_server_init()}
* mysql_server_end::            @code{mysql_server_end()}
@end menu

@node mysql_server_init, mysql_server_end, C Embedded Server func, C Embedded Server func
@subsubsection @code{mysql_server_init()}

@findex @code{mysql_server_init()}

@code{int mysql_server_init(int argc, char **argv, char **groups)}

@subsubheading Description

This function @strong{must} be called once in the program before
calling any other MySQL function.  It starts up the server and
initialises any subsystems (@code{mysys}, InnoDB, etc.) that the
server uses.  If this function is not called, the program will
crash.  If you are using the DBUG package that comes with MySQL,
you should call this after you have called @code{MY_INIT()}.

The @code{argc} and @code{argv} arguments are analogous to the
arguments to @code{main()}.  The first element of @code{argv}
is ignored (it typically contains the program name).  For
convenience, @code{argc} may be @code{0} (zero) if there are no
command-line arguments for the server.

The @code{NULL}-terminated list of strings in @code{groups}
selects which groups in the option files will be active.
@xref{Option files}.  For convenience, @code{groups} may be
@code{NULL}, in which case the @code{[server]} and @code{[emedded]} groups
will be active.

@subsubheading Example

@example
#include <mysql.h>
#include <stdlib.h>

static char *server_args[] = @{
  "this_program",       /* this string is not used */
  "--datadir=.",
  "--set-variable=key_buffer_size=32M"
@};
static char *server_groups[] = @{
  "embedded",
  "server",
  "this_program_SERVER",
  (char *)NULL
@};

int main(void) @{
  mysql_server_init(sizeof(server_args) / sizeof(char *),
                    server_args, server_groups);

  /* Use any MySQL API functions here */

  mysql_server_end();

  return EXIT_SUCCESS;
@}
@end example

@subsubheading Return Values

0 if okay, 1 if an error occurred.

@node mysql_server_end,  , mysql_server_init, C Embedded Server func
@subsubsection @code{mysql_server_end()}

@findex @code{mysql_server_end()}

@code{void mysql_server_end(void)}

@subsubheading Description

This function @strong{must} be called once in the program after
all other MySQL functions.  It shuts down the embedded server.

@subsubheading Return Values

None.

@node C API problems, Building clients, C Embedded Server func, C
@subsection Common questions and problems when using the C API

@tindex @code{mysql_query()}
@tindex @code{mysql_store_result()}

@menu
* NULL mysql_store_result::     Why Is It that After @code{mysql_query()} Returns Success, @code{mysql_store_result()} Sometimes Returns @code{NULL}?
* Query results::               What Results Can I Get From a Query?
* Getting unique ID::           How Can I Get the Unique ID for the Last Inserted Row?
* C API linking problems::      Problems Linking with the C API
@end menu


@node NULL mysql_store_result, Query results, C API problems, C API problems
@subsubsection Why Is It that After @code{mysql_query()} Returns Success, @code{mysql_store_result()} Sometimes Returns @code{NULL}?

It is possible for @code{mysql_store_result()} to return @code{NULL}
following a successful call to @code{mysql_query()}.  When this happens, it
means one of the following conditions occurred:

@itemize @bullet
@item
There was a @code{malloc()} failure (for example, if the result set was too
large).

@item
The data couldn't be read (an error occurred on the connection).

@item
The query returned no data (for example, it was an @code{INSERT},
@code{UPDATE}, or @code{DELETE}).
@end itemize

You can always check whether the statement should have produced a
non-empty result by calling @code{mysql_field_count()}.  If
@code{mysql_field_count()} returns zero, the result is empty and the last
query was a statement that does not return values (for example, an
@code{INSERT} or a @code{DELETE}).  If @code{mysql_field_count()} returns a
non-zero value, the statement should have produced a non-empty result.
See the description of the @code{mysql_field_count()} function for an
example.

You can test for an error by calling @code{mysql_error()} or
@code{mysql_errno()}.

@node Query results, Getting unique ID, NULL mysql_store_result, C API problems
@subsubsection What Results Can I Get From a Query?

In addition to the result set returned by a query, you can also get the
following information:

@itemize @bullet
@item
@code{mysql_affected_rows()} returns the number of rows affected by the last
query when doing an @code{INSERT}, @code{UPDATE}, or @code{DELETE}. An
exception is that if @code{DELETE} is used without a @code{WHERE} clause, the
table is re-created empty, which is much faster! In this case,
@code{mysql_affected_rows()} returns zero for the number of records
affected.

@item
@code{mysql_num_rows()} returns the number of rows in a result set.  With
@code{mysql_store_result()}, @code{mysql_num_rows()} may be called as soon as
@code{mysql_store_result()} returns.  With @code{mysql_use_result()},
@code{mysql_num_rows()} may be called only after you have fetched all the
rows with @code{mysql_fetch_row()}.

@item
@code{mysql_insert_id()} returns the ID generated by the last
query that inserted a row into a table with an @code{AUTO_INCREMENT} index.
@xref{mysql_insert_id, , @code{mysql_insert_id()}}.

@item
Some queries (@code{LOAD DATA INFILE ...}, @code{INSERT INTO
... SELECT ...}, @code{UPDATE}) return additional information.  The result is
returned by @code{mysql_info()}.  See the description for @code{mysql_info()}
for the format of the string that it returns.  @code{mysql_info()} returns a
@code{NULL} pointer if there is no additional information.
@end itemize


@node Getting unique ID, C API linking problems, Query results, C API problems
@subsubsection How Can I Get the Unique ID for the Last Inserted Row?

@cindex unique ID
@cindex last row, unique ID
@cindex ID, unique
@cindex tables, unique ID for last row

If you insert a record in a table containing a column that has the
@code{AUTO_INCREMENT} attribute, you can get the most recently generated
ID by calling the @code{mysql_insert_id()} function.

You can also retrieve the ID by using the @code{LAST_INSERT_ID()} function in
a query string that you pass to @code{mysql_query()}.

You can check if an @code{AUTO_INCREMENT} index is used by executing
the following code. This also checks if the query was an @code{INSERT} with
an @code{AUTO_INCREMENT} index:

@example
if (mysql_error(&mysql)[0] == 0 &&
    mysql_num_fields(result) == 0 &&
    mysql_insert_id(&mysql) != 0)
@{
    used_id = mysql_insert_id(&mysql);
@}
@end example

The most recently generated ID is maintained in the server on a
per-connection basis.  It will not be changed by another client. It will not
even be changed if you update another @code{AUTO_INCREMENT} column with a
non-magic value (that is, a value that is not @code{NULL} and not @code{0}).

If you want to use the ID that was generated for one table and insert
it into a second table, you can use SQL statements like this:

@example
INSERT INTO foo (auto,text)
    VALUES(NULL,'text');              # generate ID by inserting NULL
INSERT INTO foo2 (id,text)
    VALUES(LAST_INSERT_ID(),'text');  # use ID in second table
@end example


@node C API linking problems,  , Getting unique ID, C API problems
@subsubsection Problems Linking with the C API

@cindex linking, problems
@cindex C API, linking problems

When linking with the C API, the following errors may occur on some systems:

@example
gcc -g -o client test.o -L/usr/local/lib/mysql -lmysqlclient -lsocket -lnsl

Undefined        first referenced
 symbol          in file
floor            /usr/local/lib/mysql/libmysqlclient.a(password.o)
ld: fatal: Symbol referencing errors. No output written to client
@end example

If this happens on your system, you must include the math library by
adding @code{-lm} to the end of the compile/link line.


@node Building clients, Threaded clients, C API problems, C
@subsection Building Client Programs

@cindex client programs, building
@cindex linking
@cindex building, client programs
@cindex programs, client

If you compile MySQL clients that you've written yourself or that
you obtain from a third-party, they must be linked using the
@code{-lmysqlclient -lz} option on the link command.  You may also need to
specify a @code{-L} option to tell the linker where to find the library.  For
example, if the library is installed in @file{/usr/local/mysql/lib}, use
@code{-L/usr/local/mysql/lib -lmysqlclient -lz} on the link command.

For clients that use MySQL header files, you may need to specify a
@code{-I} option when you compile them (for example,
@code{-I/usr/local/mysql/include}), so the compiler can find the header
files.


@node Threaded clients, libmysqld, Building clients, C
@subsection How to Make a Threaded Client

@cindex clients, threaded
@cindex threaded clients

The client library is almost thread-safe. The biggest problem is
that the subroutines in @file{net.c} that read from sockets are not
interrupt safe.  This was done with the thought that you might want to
have your own alarm that can break a long read to a server.  If you
install interrupt handlers for the @code{SIGPIPE} interrupt,
the socket handling should be thread-safe.

In the older binaries we distribute on our web site
(@uref{http://www.mysql.com/}), the client libraries are not normally
compiled with the thread-safe option (the Windows binaries are by
default compiled to be thread-safe). Newer binary distributions should
have both a normal and a thread-safe client library.

To get a threaded client where you can interrupt the client from other
threads and set timeouts when talking with the MySQL server, you should
use the @code{-lmysys}, @code{-lmystrings}, and @code{-ldbug} libraries and
the @code{net_serv.o} code that the server uses.

If you don't need interrupts or timeouts, you can just compile a
thread-safe client library @code{(mysqlclient_r)} and use this.  @xref{C,,
MySQL C API}.  In this case you don't have to worry about the
@code{net_serv.o} object file or the other MySQL libraries.

When using a threaded client and you want to use timeouts and
interrupts, you can make great use of the routines in the
@file{thr_alarm.c} file.  If you are using routines from the
@code{mysys} library, the only thing you must remember is to call
@code{my_init()} first! @xref{C Thread functions}.

All functions except @code{mysql_real_connect()} are by default
thread-safe.  The following notes describe how to compile a thread-safe
client library and use it in a thread-safe manner.  (The notes below for
@code{mysql_real_connect()} actually apply to @code{mysql_connect()} as
well, but because @code{mysql_connect()} is deprecated, you should be
using @code{mysql_real_connect()} anyway.)

To make @code{mysql_real_connect()} thread-safe, you must recompile the
client library with this command:

@example
shell> ./configure --enable-thread-safe-client
@end example

This will create a thread-safe client library @code{libmysqlclient_r}.
(Assuming your OS has a thread-safe @code{gethostbyname_r()} function.)
This library is thread-safe per connection.  You can let two threads
share the same connection with the following caveats:

@itemize @bullet
@item
Two threads can't send a query to the MySQL server at the same time on
the same connection.  In particular, you have to ensure that between a
@code{mysql_query()} and @code{mysql_store_result()} no other thread is using
the same connection.
@item
Many threads can access different result sets that are retrieved with
@code{mysql_store_result()}.
@item
If you use @code{mysql_use_result}, you have to ensure that no other thread
is using the same connection until the result set is closed.
However, it really is best for threaded clients that share the same
connection to use @code{mysql_store_result()}.
@item
If you want to use multiple threads on the same connection, you must
have a mutex lock around your @code{mysql_query()} and
@code{mysql_store_result()} call combination.  Once
@code{mysql_store_result()} is ready, the lock can be released and other
threads may query the same connection.
@item
If you program with POSIX threads, you can use
@code{pthread_mutex_lock()} and @code{pthread_mutex_unlock()} to
establish and release a mutex lock.
@end itemize

You need to know the following if you have a thread that is calling
MySQL functions which did not create the connection to the
MySQL database:

When you call @code{mysql_init()} or @code{mysql_connect()}, MySQL will
create a thread specific variable for the thread that is used by the
debug library (among other things).

If you call a MySQL function, before the thread has
called @code{mysql_init()} or @code{mysql_connect()}, the thread will
not have the necessary thread specific variables in place and you are
likely to end up with a core dump sooner or later.

The get things to work smoothly you have to do the following:

@enumerate
@item
Call @code{my_init()} at the start of your program if it calls
any other MySQL function before calling @code{mysql_real_connect()}.
@item
Call @code{mysql_thread_init()} in the thread handler before calling
any MySQL function.
@item
In the thread, call @code{mysql_thread_end()} before calling
@code{pthread_exit()}.  This will free the memory used by MySQL thread
specific variables.
@end enumerate

You may get some errors because of undefined symbols when linking your
client with @code{libmysqlclient_r}. In most cases this is because you haven't
included the thread libraries on the link/compile line.

@node libmysqld,  , Threaded clients, C
@subsection libmysqld, the Embedded MySQL Server Library

@cindex libmysqld
@cindex embedded MySQL server library

@menu
* libmysqld overview::          Overview of the Embedded MySQL Server Library
* libmysqld compiling::         Compiling Programs with @code{libmysqld}
* libmysqld restrictions::      Restrictions when Using the Embedded MySQL Server
* libmysqld options::           Using Option Files with the Embedded Server
* libmysqld TODO::              Things left to do in Embedded Server (TODO)
* libmysqld example::           A Simple Embedded Server Example
* libmysqld licensing::         Licensing the Embedded Server
@end menu

@node libmysqld overview, libmysqld compiling, libmysqld, libmysqld
@subsubsection Overview of the Embedded MySQL Server Library

The embedded MySQL server library makes it possible to run a
full-featured MySQL server inside the client application.  The
main benefits are increased speed and more simple management for
embedded applications.

The API is identical for the embedded MySQL version and the
client/server version.  To change an old threaded application to use the
embedded library, you normally only have to add calls to the following
functions:

@multitable @columnfractions .25 .70
@item @strong{Function} @tab @strong{When to call}
@item @code{mysql_server_init()} @tab Should be called before any other MySQL function is called, preferably early in the @code{main()} function.
@item @code{mysql_server_end()} @tab Should be called before your program exits.
@item @code{mysql_thread_init()} @tab Should be called in each thread you create that will access MySQL.
@item @code{mysql_thread_end()} @tab Should be called before calling @code{pthread_exit()}
@end multitable

Then you must link your code with @code{libmysqld.a} instead of @code{libmysqlclient.a}.

The above @code{mysql_server_xxx} functions are also included in
@code{libmysqlclient.a} to allow you to change between the embedded and the
client/server version by just linking your application with the right
library. @xref{mysql_server_init}.


@node libmysqld compiling, libmysqld restrictions, libmysqld overview, libmysqld
@subsubsection Compiling Programs with @code{libmysqld}

To get a @code{libmysqld} library you should configure MySQL with the
@code{--with-embedded-server} option.

When you link your program with @code{libmysqld}, you must also include
the system-specific @code{pthread} libraries and some libraries that
the MySQL server uses.  You can get the full list of libraries by executing
@code{mysql_config --libmysqld-libs}.

The correct flags for compiling and linking a threaded program
must be used, even if you do not directly call any thread
functions in your code.

@node libmysqld restrictions, libmysqld options, libmysqld compiling, libmysqld
@subsubsection Restrictions when using the Embedded MySQL Server

The embedded server has the following limitations:

@itemize @bullet
@item
No support for ISAM tables. (This is mainly done to make the library smaller)
@item
No UDF functions.
@item
No stack trace on core dump.
@item
No internal RAID support. (This is not normally needed as most OS has
nowadays support for big files).
@item
You can set this up as a server or a master (no replication).
@item
You can't connect to the embedded server from an outside process with
sockets or TCP/IP.
@end itemize

Some of these limitations can be changed by editing the @file{mysql_embed.h}
include file and recompiling MySQL.

@node libmysqld options, libmysqld TODO, libmysqld restrictions, libmysqld
@subsubsection Using Option Files with the Embedded Server

@cindex defaults, embedded

The following is the recommended way to use option files to make it easy
to switch between a client/server application and one where MySQL is
embedded. @xref{Option files}.

@itemize @bullet
@item
Put common options in the @code{[server]} section. These will be read by
both MySQL versions.
@item
Put client/server specific options in the @code{[mysqld]} section.
@item
Put embedded MySQL specific options in the @code{[embedded]} section.
@item
Put application specific options in a @code{[ApplicationName_SERVER]}
section.
@end itemize

@node libmysqld TODO, libmysqld example, libmysqld options, libmysqld
@subsubsection Things left to do in Embedded Server (TODO)

@cindex TODO, embedded server

@itemize @bullet
@item
Currently we only provide a static version of the @code{mysqld} library,
in the future we will also provide a shared library for this.
@item
We are going to provide options to leave out some parts of MySQL to make
the library smaller.
@item
There is still a lot of speed optimisation to do.
@item
Errors are written to stderr.  We will add an option to specify a
filename for these.
@item
We have to change InnoDB to not be so verbose when using in the embedded
version.
@end itemize


@node libmysqld example, libmysqld licensing, libmysqld TODO, libmysqld
@subsubsection A Simple Embedded Server Example

This example program and makefile should work without any
changes on a Linux or FreeBSD system.  For other operating
systems, minor changes will be needed.  This example is
designed to give enough details to understand the problem,
without the clutter that is a necessary part of a real
application.

To try out the example, create an @file{test_libmysqld} directory
at the same level as the mysql-4.0 source directory.  Save
the @file{test_libmysqld.c} source and the @file{GNUmakefile} in the
directory, and run GNU @file{make} from inside the @file{test_libmysqld}
directory.

@file{test_libmysqld.c}
@example
/*
 * A simple example client, using the embedded MySQL server library
 */

#include <mysql.h>
#include <stdarg.h>
#include <stdio.h>
#include <stdlib.h>

MYSQL *db_connect(const char *dbname);
void db_disconnect(MYSQL *db);
void db_do_query(MYSQL *db, const char *query);

const char *server_groups[] = @{
  "test_libmysqld_SERVER", "embedded", "server", NULL
@};

int
main(int argc, char **argv)
@{
  MYSQL *one, *two;

  /* mysql_server_init() must be called before any other mysql
   * functions.
   *
   * You can use mysql_server_init(0, NULL, NULL), and it will
   * initialise the server using groups = @{
   *   "server", "embedded", NULL
   *  @}.
   *
   * In your $HOME/.my.cnf file, you probably want to put:

[test_libmysqld_SERVER]
language = /path/to/source/of/mysql/sql/share/english

   * You could, of course, modify argc and argv before passing
   * them to this function.  Or you could create new ones in any
   * way you like.  But all of the arguments in argv (except for
   * argv[0], which is the program name) should be valid options
   * for the MySQL server.
   *
   * If you link this client against the normal mysqlclient
   * library, this function is just a stub that does nothing.
   */
  mysql_server_init(argc, argv, (char **)server_groups);

  one = db_connect("test");
  two = db_connect(NULL);

  db_do_query(one, "SHOW TABLE STATUS");
  db_do_query(two, "SHOW DATABASES");

  mysql_close(two);
  mysql_close(one);

  /* This must be called after all other mysql functions */
  mysql_server_end();

  exit(EXIT_SUCCESS);
@}

static void
die(MYSQL *db, char *fmt, ...)
@{
  va_list ap;
  va_start(ap, fmt);
  vfprintf(stderr, fmt, ap);
  va_end(ap);
  (void)putc('\n', stderr);
  if (db)
    db_disconnect(db);
  exit(EXIT_FAILURE);
@}

MYSQL *
db_connect(const char *dbname)
@{
  MYSQL *db = mysql_init(NULL);
  if (!db)
    die(db, "mysql_init failed: no memory");
  /*
   * Notice that the client and server use separate group names.
   * This is critical, because the server will not accept the
   * client's options, and vice versa.
   */
  mysql_options(db, MYSQL_READ_DEFAULT_GROUP, "test_libmysqld_CLIENT");
  if (!mysql_real_connect(db, NULL, NULL, NULL, dbname, 0, NULL, 0))
    die(db, "mysql_real_connect failed: %s", mysql_error(db));

  return db;
@}

void
db_disconnect(MYSQL *db)
@{
  mysql_close(db);
@}

void
db_do_query(MYSQL *db, const char *query)
@{
  if (mysql_query(db, query) != 0)
    goto err;

  if (mysql_field_count(db) > 0)
  @{
    MYSQL_RES   *res;
    MYSQL_ROW    row, end_row;
    int num_fields;

    if (!(res = mysql_store_result(db)))
      goto err;
    num_fields = mysql_num_fields(res);
    while ((row = mysql_fetch_row(res)))
    @{
      (void)fputs(">> ", stdout);
      for (end_row = row + num_fields; row < end_row; ++row)
        (void)printf("%s\t", row ? (char*)*row : "NULL");
      (void)fputc('\n', stdout);
    @}
    (void)fputc('\n', stdout);
  @}
  else
    (void)printf("Affected rows: %lld\n", mysql_affected_rows(db));

  return;

err:
  die(db, "db_do_query failed: %s [%s]", mysql_error(db), query);
@}
@end example

@file{GNUmakefile}
@example
# This assumes the MySQL software is installed in /usr/local/mysql
inc      := /usr/local/mysql/include/mysql
lib      := /usr/local/mysql/lib

# If you have not installed the MySQL software yet, try this instead
#inc      := $(HOME)/mysql-4.0/include
#lib      := $(HOME)/mysql-4.0/libmysqld

CC       := gcc
CPPFLAGS := -I$(inc) -D_THREAD_SAFE -D_REENTRANT
CFLAGS   := -g -W -Wall
LDFLAGS  := -static
# You can change -lmysqld to -lmysqlclient to use the
# client/server library
LDLIBS    = -L$(lib) -lmysqld -lz -lm -lcrypt

ifneq (,$(shell grep FreeBSD /COPYRIGHT 2>/dev/null))
# FreeBSD
LDFLAGS += -pthread
else
# Assume Linux
LDLIBS += -lpthread
endif

# This works for simple one-file test programs
sources := $(wildcard *.c)
objects := $(patsubst %c,%o,$(sources))
targets := $(basename $(sources))

all: $(targets)

clean:
	rm -f $(targets) $(objects) *.core
@end example

@node libmysqld licensing,  , libmysqld example, libmysqld
@subsubsection Licensing the Embedded Server

The MySQL source code is covered by the GNU GPL license
(@pxref{GPL license}).  One result of this is that any program
which includes, by linking with @code{libmysqld}, the MySQL
source code must be released as free software (under a license
compatible with the GPL).

We encourage everyone to promote free software by releasing
code under the GPL or a compatible license.  For those who
are not able to do this, another option is to purchase a
commercial licence for the MySQL code from MySQL AB.
For details, please see @ref{MySQL licenses}.

@node Cplusplus, Java, C, Clients
@section MySQL C++ APIs

@menu
* Borland C++::                 Borland C++
@end menu


@cindex C++ APIs

Two APIs are available in the MySQL Contrib directory
(@uref{http://www.mysql.com/Downloads/Contrib/}).


@node Borland C++,  , Cplusplus, Cplusplus
@subsection Borland C++

@cindex Borland C++ compiler

You can compile the MySQL Windows source with Borland C++ 5.02.
(The Windows source includes only projects for Microsoft VC++, for
Borland C++ you have to do the project files yourself.)

One known problem with Borland C++ is that it uses a different structure
alignment than VC++.  This means that you will run into problems if you
try to use the default @code{libmysql.dll} libraries (that was compiled
with VC++) with Borland C++. You can do one of the following to avoid
this problem.

@itemize @bullet
@item
You can use the static MySQL libraries for Borland C++ that you
can find on @uref{http://www.mysql.com/downloads/os-win32.html}.
@item
Only call @code{mysql_init()} with @code{NULL} as an argument, not a
pre-allocated MYSQL struct.
@end itemize


@node Java, Python, Cplusplus, Clients
@section MySQL Java Connectivity (JDBC)

@cindex Java connectivity
@cindex JDBC

There are 2 supported JDBC drivers for MySQL (the mm driver and
the Reisin JDBC driver).  You can find a copy of the mm driver at
@uref{http://mmmysql.sourceforge.net/} or
@uref{http://www.mysql.com/Downloads/Contrib/} and the Reisin driver at
@uref{http://www.caucho.com/projects/jdbc-mysql/index.xtp} For
documentation consult any JDBC documentation and the driver's own
documentation for MySQL-specific features.


@node Python, Tcl, Java, Clients
@section MySQL Python APIs

@cindex Python APIs

The MySQL Contrib directory
(@uref{http://www.mysql.com/Downloads/Contrib/})
contains a Python interface written by Joseph Skinner.


@node Tcl, Eiffel, Python, Clients
@section MySQL Tcl APIs

@cindex Tcl APIs

@uref{http://www.binevolve.com/~tdarugar/tcl-sql/} (Tcl at binevolve).
The Contrib directory
(@uref{http://www.mysql.com/Downloads/Contrib/}) contains a Tcl
interface that is based on msqltcl 1.50.


@node Eiffel,  , Tcl, Clients
@section MySQL Eiffel wrapper

@cindex Eiffel Wrapper
@cindex wrappers, Eiffel

The MySQL Contrib directory
(@uref{http://www.mysql.com/Downloads/Contrib/})
contains an Eiffel wrapper written by Michael Ravits.


@node Extending MySQL, Problems, Clients, Top
@chapter Extending MySQL

@menu
* MySQL internals::             MySQL Internals
* Adding functions::            Adding New Functions to MySQL
* Adding procedures::           Adding New Procedures to MySQL
@end menu


@node MySQL internals, Adding functions, Extending MySQL, Extending MySQL
@section MySQL Internals

@cindex internals
@cindex threads

This chapter describes a lot of things that you need to know when
working on the MySQL code. If you plan to contribute to MySQL
development, want to have access to the bleeding-edge in-between
versions code, or just want to keep track of development, follow the
instructions in @ref{Installing source tree}.
If you are interested in MySQL internals, you should also subscribe
to our @code{internals} mailing list. This list is relatively low
traffic. For details on how to subscribe, please see
@ref{Mailing-list}.
All developers at MySQL AB are on the @code{internals} list and we
help other people who are working on the MySQL code. Feel free to
use this list both to ask questions about the code and to send
patches that you would like to contribute to the MySQL project!

@menu
* MySQL threads::               MySQL threads
* MySQL test suite::            MySQL test suite
@end menu


@node MySQL threads, MySQL test suite, MySQL internals, MySQL internals
@subsection MySQL Threads

The MySQL server creates the following threads:

@itemize @bullet

@item
The TCP/IP connection thread handles all connection requests and
creates a new dedicated thread to handle the authentication and
SQL query processing for each connection.

@item
On Windows NT there is a named pipe handler thread that does the same work as
the TCP/IP connection thread on named pipe connect requests.

@item
The signal thread handles all signals.  This thread also normally handles
alarms and calls @code{process_alarm()} to force timeouts on connections
that have been idle too long.

@item
If @code{mysqld} is compiled with @code{-DUSE_ALARM_THREAD}, a dedicated
thread that handles alarms is created.  This is only used on some systems where
there are problems with @code{sigwait()} or if one wants to use the
@code{thr_alarm()} code in ones application without a dedicated signal
handling thread.

@item
If one uses the @code{--flush_time=#} option, a dedicated thread is created
to flush all tables at the given interval.

@item
Every connection has its own thread.

@item
Every different table on which one uses @code{INSERT DELAYED} gets its
own thread.

@item
If you use @code{--master-host}, a slave replication thread will be
started to read and apply updates from the master.
@end itemize

@code{mysqladmin processlist} only shows the connection, @code{INSERT DELAYED},
and replication threads.


@node MySQL test suite,  , MySQL threads, MySQL internals
@subsection MySQL Test Suite

@cindex mysqltest, MySQL Test Suite
@cindex testing mysqld, mysqltest

Until recently, our main full-coverage test suite was based on proprietary
customer data and for that reason has not been publicly available. The only
publicly available part of our testing process consisted of the @code{crash-me}
test, a Perl DBI/DBD benchmark found in the @code{sql-bench} directory, and
miscellaneous tests located in @code{tests} directory. The lack of a
standardised publicly available test suite has made it difficult for our users,
as well developers, to do regression tests on the MySQL code. To
address this problem, we have created a new test system that is included in
the source and binary distributions starting in Version 3.23.29.

The current set of test cases doesn't test everything in MySQL, but it
should catch most obvious bugs in the SQL processing code, OS/library
issues, and is quite thorough in testing replication.  Our eventual goal
is to have the tests cover 100% of the code.  We welcome contributions
to our test suite.  You may especially want to contribute tests that
examine the functionality critical to your system, as this will ensure
that all future MySQL releases will work well with your
applications.

@menu
* running mysqltest::           Running the MySQL Test Suite
* extending mysqltest::         Extending the MySQL Test Suite
* Reporting mysqltest bugs::    Reporting Bugs in the MySQL Test Suite
@end menu


@node running mysqltest, extending mysqltest, MySQL test suite, MySQL test suite
@subsubsection Running the MySQL Test Suite

The test system consist of a test language interpreter
(@code{mysqltest}), a shell script to run all
tests(@code{mysql-test-run}), the actual test cases written in a special
test language, and their expected results.  To run the test suite on
your system after a build, type @code{make test} or
@code{mysql-test/mysql-test-run} from the source root.  If you have
installed a binary distribution, @code{cd} to the install root
(eg. @code{/usr/local/mysql}), and do @code{scripts/mysql-test-run}.
All tests should succeed.  If not, you should try to find out why and
report the problem if this is a bug in MySQL.
@xref{Reporting mysqltest bugs}.

If you have a copy of @code{mysqld} running on the machine where you want to
run the test suite you do not have to stop it, as long as it is not using
ports @code{9306} and @code{9307}.  If one of those ports is taken, you should
edit @code{mysql-test-run} and change the values of the master and/or slave
port to one that is available.

You can run one individual test case with
@code{mysql-test/mysql-test-run test_name}.

If one test fails, you should test running @code{mysql-test-run} with
the @code{--force} option to check if any other tests fails.


@node extending mysqltest, Reporting mysqltest bugs, running mysqltest, MySQL test suite
@subsubsection Extending the MySQL Test Suite

You can use the @code{mysqltest} language to write your own test cases.
Unfortunately, we have not yet written full documentation for it - we plan to
do this shortly.  You can, however, look at our current test cases and use
them as an example.  The following points should help you get started:

@itemize @bullet
@item
The tests are located in @code{mysql-test/t/*.test}

@item
A test case consists of @code{;} terminated statements and is similar to the
input of @code{mysql} command-line client.  A statement by default is a query
to be sent to MySQL server, unless it is recognised as internal
command (eg. @code{sleep}).

@item
All queries that produce results@-e.g., @code{SELECT}, @code{SHOW},
@code{EXPLAIN}, etc., must be preceded with @code{@@/path/to/result/file}.  The
file must contain the expected results.  An easy way to generate the result
file is to run @code{mysqltest -r < t/test-case-name.test} from
@code{mysql-test} directory, and then edit the generated result files, if
needed, to adjust them to the expected output.  In that case, be very careful
about not adding or deleting any invisible characters - make sure to only
change the text and/or delete lines.  If you have to insert a line, make sure
the fields are separated with a hard tab, and there is a hard tab at the end.
You may want to use @code{od -c} to make sure your text editor has not messed
anything up during edit.  We, of course, hope that you will never have to edit
the output of @code{mysqltest -r} as you only have to do it when you find a
bug.

@item
To be consistent with our setup, you should put your result files in
@code{mysql-test/r} directory and name them @code{test_name.result}.  If the
test produces more than one result, you should use @code{test_name.a.result},
@code{test_name.b.result}, etc.

@item
If a statement returns an error, you should on the line before the statement
specify with the @code{--error error-number}.  The error number can be
a list of possible error numbers separated with @code{','}.

@item
If you are writing a replication test case, you should on the first line of
the test file, put @code{source include/master-slave.inc;}.  To switch between
master and slave, use @code{connection master;} and @code{connection slave;}.
If you need to do something on an alternate connection, you can do
@code{connection master1;} for the master, and @code{connection slave1;} for
the slave.

@item
If you need to do something in a loop, you can use something like this:
@example
let $1=1000;
while ($1)
@{
 # do your queries here
 dec $1;
@}
@end example

@item
To sleep between queries, use the @code{sleep} command. It supports fractions
of a second, so you can do @code{sleep 1.3;}, for example, to sleep 1.3
seconds.

@item
To run the slave with additional options for your test case, put them
in the command-line format in @code{mysql-test/t/test_name-slave.opt}. For
the master, put them in @code{mysql-test/t/test_name-master.opt}.

@item
If you have a question about the test suite, or have a test case to contribute,
e-mail to @email{internals@@lists.mysql.com}. As the list does not accept
attachments, you should ftp all the relevant files to:
@uref{ftp://support.mysql.com/pub/mysql/Incoming/}

@end itemize


@node Reporting mysqltest bugs,  , extending mysqltest, MySQL test suite
@subsubsection Reporting Bugs in the MySQL Test Suite

If your MySQL version doesn't pass the test suite you should
do the following:

@itemize @bullet
@item
Don't send a bug report before you have found out as much as possible of
what when wrong!  When you do it, please use the @code{mysqlbug} script
so that we can get information about your system and @code{MySQL}
version. @xref{Bug reports}.
@item
Make sure to include the output of @code{mysql-test-run}, as well as
contents of all @code{.reject} files in @code{mysql-test/r} directory.
@item
If a test in the test suite fails, check if the test fails also when run
by its own:

@example
cd mysql-test
mysql-test-run --local test-name
@end example

If this fails, then you should configure MySQL with
@code{--with-debug} and run @code{mysql-test-run} with the
@code{--debug} option. If this also fails send the trace file
@file{var/tmp/master.trace} to ftp://support.mysql.com/pub/mysql/secret
so that we can examine it. Please remember to also include a full
description of your system, the version of the mysqld binary and how you
compiled it.

@item
Try also to run @code{mysql-test-run} with the @code{--force} option to
see if there is any other test that fails.

@item
If you have compiled MySQL yourself, check our manual for how
to compile MySQL on your platform or, preferable, use one of
the binaries we have compiled for you at
@uref{http://www.mysql.com/downloads/}.  All our standard binaries should
pass the test suite !

@item
If you get an error, like @code{Result length mismatch} or @code{Result
content mismatch} it means that the output of the test didn't match
exactly the expected output. This could be a bug in MySQL or
that your mysqld version produces slight different results under some
circumstances.

Failed test results are put in a file with the same base name as the
result file with the @code{.reject} extension.  If your test case is
failing, you should do a diff on the two files.  If you cannot see how
they are different, examine both with @code{od -c} and also check their
lengths.

@item
If a test fails totally, you should check the logs file in the
@code{mysql-test/var/log} directory for hints of what went wrong.

@item
If you have compiled MySQL with debugging you can try to debug this
by running @code{mysql-test-run} with the @code{--gdb} and/or @code{--debug}
options.
@xref{Making trace files}.

If you have not compiled MySQL for debugging you should probably
do that.  Just specify the @code{--with-debug} options to @code{configure}!
@xref{Installing source}.
@end itemize


@node Adding functions, Adding procedures, MySQL internals, Extending MySQL
@section Adding New Functions to MySQL

@cindex functions, new
@cindex adding, new functions
@cindex user-defined functions, adding
@cindex UDFs, defined
@cindex functions, user-defined

There are two ways to add new functions to MySQL:

@itemize @bullet
@item You can add the function through the user-definable function (UDF)
interface.  User-definable functions are added and removed dynamically using
the @code{CREATE FUNCTION} and @code{DROP FUNCTION} statements.
@xref{CREATE FUNCTION, , @code{CREATE FUNCTION}}.

@item You can add the function as a native (built in) MySQL function.
Native functions are compiled into the @code{mysqld} server and become
available on a permanent basis.
@end itemize

Each method has advantages and disadvantages:

@itemize @bullet
@item
If you write a user-definable function, you must install the object file
in addition to the server itself.  If you compile your function into the
server, you don't need to do that.
@item
You can add UDFs to a binary MySQL distribution.  Native functions
require you to modify a source distribution.
@item
If you upgrade your MySQL distribution, you can continue to use your
previously installed UDFs.  For native functions, you must repeat your
modifications each time you upgrade.
@end itemize

Whichever method you use to add new functions, they may be used just like
native functions such as @code{ABS()} or @code{SOUNDEX()}.

@menu
* CREATE FUNCTION::             @code{CREATE FUNCTION/DROP FUNCTION} Syntax
* Adding UDF::                  Adding a new user-definable function
* Adding native function::      Adding a new native function
@end menu


@node CREATE FUNCTION, Adding UDF, Adding functions, Adding functions
@subsection @code{CREATE FUNCTION/DROP FUNCTION} Syntax

@findex CREATE FUNCTION
@findex DROP FUNCTION
@findex UDF functions
@findex User-defined functions
@findex Functions, user-defined

@example
CREATE [AGGREGATE] FUNCTION function_name RETURNS @{STRING|REAL|INTEGER@}
       SONAME shared_library_name

DROP FUNCTION function_name
@end example

A user-definable function (UDF) is a way to extend MySQL with a new
function that works like native (built in) MySQL functions such as
@code{ABS()} and @code{CONCAT()}.

@code{AGGREGATE} is a new option for MySQL Version 3.23.  An
@code{AGGREGATE} function works exactly like a native MySQL
@code{GROUP} function like @code{SUM} or @code{COUNT()}.

@code{CREATE FUNCTION} saves the function's name, type, and shared library
name in the @code{mysql.func} system table.  You must have the
@strong{insert} and @strong{delete} privileges for the @code{mysql} database
to create and drop functions.

All active functions are reloaded each time the server starts, unless
you start @code{mysqld} with the @code{--skip-grant-tables} option.  In
this case, UDF initialisation is skipped and UDFs are unavailable.
(An active function is one that has been loaded with @code{CREATE FUNCTION}
and not removed with @code{DROP FUNCTION}.)

For instructions on writing user-definable functions, see @ref{Adding
functions}.  For the UDF mechanism to work, functions must be written in C or
C++, your operating system must support dynamic loading and you must have
compiled @code{mysqld} dynamically (not statically).

Note that to make @code{AGGREGATE} work, you must have a
@code{mysql.func} table that contains the column @code{type}.  If this
is not the case, you should run the script
@code{mysql_fix_privilege_tables} to get this fixed.


@node Adding UDF, Adding native function, CREATE FUNCTION, Adding functions
@subsection Adding a New User-definable Function

@cindex adding, user-definable functions
@cindex user-defined functions, adding
@cindex functions, user-definable, adding

@menu
* UDF calling::                 UDF Calling Sequences
* UDF aggr. calling ::          UDF Calling Sequences for aggregate functions
* UDF arguments::               Argument Processing
* UDF return values::           Return Values and Error Handling
* UDF compiling::               Compiling and Installing User-definable Functions
@end menu


For the UDF mechanism to work, functions must be written in C or C++ and your
operating system must support dynamic loading.  The MySQL source
distribution includes a file @file{sql/udf_example.cc} that defines 5 new
functions.  Consult this file to see how UDF calling conventions work.

For @code{mysqld} to be able to use UDF functions, you should configure MySQL
with @code{--with-mysqld-ldflags=-rdynamic} The reason is that to on
many platforms (including Linux) you can load a dynamic library (with
@code{dlopen()}) from a static linked program, which you would get if
you are using @code{--with-mysqld-ldflags=-all-static} If you want to
use an UDF that needs to access symbols from @code{mysqld} (like the
@code{methaphone} example in @file{sql/udf_example.cc} that uses
@code{default_charset_info}), you must link the program with
@code{-rdynamic} (see @code{man dlopen}).

For each function that you want to use in SQL statements, you should define
corresponding C (or C++) functions.  In the discussion below, the name
``xxx'' is used for an example function name.  To distinquish between SQL and
C/C++ usage, @code{XXX()} (uppercase) indicates a SQL function call, and
@code{xxx()} (lowercase) indicates a C/C++ function call.

The C/C++ functions that you write to implement the interface for
@code{XXX()} are:

@table @asis
@item @code{xxx()} (required)
The main function.  This is where the function result is computed.
The correspondence between the SQL type and return type of your C/C++
function is shown here:

@multitable @columnfractions .15 .15
@item @strong{SQL type} @tab @strong{C/C++ type}
@item @code{STRING}     @tab @code{char *}
@item @code{INTEGER}    @tab @code{long long}
@item @code{REAL}       @tab @code{double}
@end multitable

@item @code{xxx_init()} (optional)
The initialisation function for @code{xxx()}.  It can be used to:

@itemize @bullet
@item
Check the number of arguments to @code{XXX()}.
@item
Check that the arguments are of a required type or, alternatively,
tell MySQL to coerce arguments to the types you want when
the main function is called.
@item
Allocate any memory required by the main function.
@item
Specify the maximum length of the result.
@item
Specify (for @code{REAL} functions) the maximum number of decimals.
@item
Specify whether the result can be @code{NULL}.
@end itemize

@item @code{xxx_deinit()} (optional)
The deinitialisation function for @code{xxx()}.  It should deallocate any
memory allocated by the initialisation function.
@end table

When a SQL statement invokes @code{XXX()}, MySQL calls the
initialisation function @code{xxx_init()} to let it perform any required
setup, such as argument checking or memory allocation.  If @code{xxx_init()}
returns an error, the SQL statement is aborted with an error message and the
main and deinitialisation functions are not called.  Otherwise, the main
function @code{xxx()} is called once for each row.  After all rows have been
processed, the deinitialisation function @code{xxx_deinit()} is called so it
can perform any required cleanup.

For aggregate functions (like @code{SUM()}), you must also provide the
following functions:

@table @asis
@item @code{xxx_reset()} (required)
Reset sum and insert the argument as the initial value for a new group.
@item @code{xxx_add()} (required)
Add the argument to the old sum.
@end table

When using aggregate UDF functions MySQL works the following way:

@enumerate
@item
Call @code{xxx_init()} to let the aggregate function allocate the memory it
will need to store results.
@item
Sort the table according to the @code{GROUP BY} expression.
@item
For the first row in a new group, call the @code{xxx_reset()} function.
@item
For each new row that belongs in the same group, call the
@code{xxx_add()} function.
@item
When the group changes or after the last row has been processed,
call @code{xxx()} to get the result for the aggregate.
@item
Repeat 3-5 until all rows has been processed
@item
Call @code{xxx_deinit()} to let the UDF free any memory it has allocated.
@end enumerate

All functions must be thread-safe (not just the main function,
but the initialisation and deinitialisation functions as well). This means
that you are not allowed to allocate any global or static variables that
change!  If you need memory, you should allocate it in @code{xxx_init()}
and free it in @code{xxx_deinit()}.


@node UDF calling, UDF aggr. calling , Adding UDF, Adding UDF
@subsubsection UDF Calling Sequences for simple functions

@cindex calling sequences for simple functions, UDF

The main function should be declared as shown here.  Note that the return
type and parameters differ, depending on whether you will declare the SQL
function @code{XXX()} to return @code{STRING}, @code{INTEGER}, or @code{REAL}
in the @code{CREATE FUNCTION} statement:

@noindent
For @code{STRING} functions:

@example
char *xxx(UDF_INIT *initid, UDF_ARGS *args,
          char *result, unsigned long *length,
          char *is_null, char *error);
@end example

@noindent
For @code{INTEGER} functions:

@example
long long xxx(UDF_INIT *initid, UDF_ARGS *args,
              char *is_null, char *error);
@end example

@noindent
For @code{REAL} functions:

@example
double xxx(UDF_INIT *initid, UDF_ARGS *args,
              char *is_null, char *error);
@end example

The initialisation and deinitialisation functions are declared like this:

@example
my_bool xxx_init(UDF_INIT *initid, UDF_ARGS *args, char *message);

void xxx_deinit(UDF_INIT *initid);
@end example

The @code{initid} parameter is passed to all three functions.  It points to a
@code{UDF_INIT} structure that is used to communicate information between
functions.  The @code{UDF_INIT} structure members are listed below.  The
initialisation function should fill in any members that it wishes to change.
(To use the default for a member, leave it unchanged.):

@table @code
@item my_bool maybe_null
@code{xxx_init()} should set @code{maybe_null} to @code{1} if @code{xxx()}
can return @code{NULL}.  The default value is @code{1} if any of the
arguments are declared @code{maybe_null}.

@item unsigned int decimals
Number of decimals.  The default value is the maximum number of decimals in
the arguments passed to the main function.  (For example, if the function is
passed @code{1.34}, @code{1.345}, and @code{1.3}, the default would be 3,
because @code{1.345} has 3 decimals.

@item unsigned int max_length
The maximum length of the string result.  The default value differs depending
on the result type of the function.  For string functions, the default is the
length of the longest argument.  For integer functions, the default is 21
digits.  For real functions, the default is 13 plus the number of decimals
indicated by @code{initid->decimals}.  (For numeric functions, the length
includes any sign or decimal point characters.)

If you want to return a blob, you can set this to 65K or 16M; this
memory is not allocated but used to decide which column type to use if
there is a need to temporary store the data.

@item char *ptr
A pointer that the function can use for its own purposes.  For example,
functions can use @code{initid->ptr} to communicate allocated memory
between functions.  In @code{xxx_init()}, allocate the memory and assign it
to this pointer:

@example
initid->ptr = allocated_memory;
@end example

In @code{xxx()} and @code{xxx_deinit()}, refer to @code{initid->ptr} to use
or deallocate the memory.
@end table


@node UDF aggr. calling , UDF arguments, UDF calling, Adding UDF
@subsubsection UDF Calling Sequences for aggregate functions

@cindex calling sequences for aggregate functions, UDF

Here follows a description of the different functions you need to define
when you want to create an aggregate UDF function.

@example
char *xxx_reset(UDF_INIT *initid, UDF_ARGS *args,
                char *is_null, char *error);
@end example

This function is called when MySQL finds the first row in a new group.
In the function you should reset any internal summary variables and then set
the given argument as the first argument in the group.

In many cases this is implemented internally by reseting all variables
and then calling @code{xxx_add()}.

@example
char *xxx_add(UDF_INIT *initid, UDF_ARGS *args,
              char *is_null, char *error);
@end example

This function is called for all rows that belongs to the same group,
except for the first row. In this you should add the value in UDF_ARGS
to your internal summary variable.

The @code{xxx()} function should be declared identical as when you
define a simple UDF function. @xref{UDF calling}.

This function is called when all rows in the group has been processed.
You should normally never access the @code{args} variable here but
return your value based on your internal summary variables.

All argument processing in @code{xxx_reset()} and @code{xxx_add()}
should be done identically as for normal UDF functions. @xref{UDF
arguments}.

The return value handling in @code{xxx()} should be done identically as
for a normal UDF. @xref{UDF return values}.

The pointer argument to @code{is_null} and @code{error} is the same for
all calls to @code{xxx_reset()}, @code{xxx_add()} and @code{xxx()}.
You can use this to remember that you got an error or if the @code{xxx()}
function should return @code{NULL}. Note that you should not store a string
into @code{*error}!  This is just a 1 byte flag!

@code{is_null} is reset for each group (before calling @code{xxx_reset()}.
@code{error} is never reset.

If @code{isnull} or @code{error} are set after @code{xxx()} then MySQL
will return @code{NULL} as the result for the group function.

@node UDF arguments, UDF return values, UDF aggr. calling , Adding UDF
@subsubsection Argument Processing

@cindex argument processing
@cindex processing, arguments

The @code{args} parameter points to a @code{UDF_ARGS} structure that has the
members listed here:

@table @code
@item unsigned int arg_count
The number of arguments.  Check this value in the initialisation function
if you want your function to be called with a particular number of arguments.
For example:

@example
if (args->arg_count != 2)
@{
    strcpy(message,"XXX() requires two arguments");
    return 1;
@}
@end example


@item enum Item_result *arg_type
The types for each argument.  The possible type values are
@code{STRING_RESULT}, @code{INT_RESULT}, and @code{REAL_RESULT}.

To make sure that arguments are of a given type and return an
error if they are not, check the @code{arg_type} array in the initialisation
function.  For example:

@example
if (args->arg_type[0] != STRING_RESULT ||
    args->arg_type[1] != INT_RESULT)
@{
    strcpy(message,"XXX() requires a string and an integer");
    return 1;
@}
@end example

As an alternative to requiring your function's arguments to be of particular
types, you can use the initialisation function to set the @code{arg_type}
elements to the types you want.  This causes MySQL to coerce
arguments to those types for each call to @code{xxx()}.  For example, to
specify coercion of the first two arguments to string and integer, do this in
@code{xxx_init()}:

@example
args->arg_type[0] = STRING_RESULT;
args->arg_type[1] = INT_RESULT;
@end example

@item char **args
@code{args->args} communicates information to the initialisation function
about the general nature of the arguments your function was called with.  For a
constant argument @code{i}, @code{args->args[i]} points to the argument
value.  (See below for instructions on how to access the value properly.)
For a non-constant argument, @code{args->args[i]} is @code{0}.
A constant argument is an expression that uses only constants, such as
@code{3} or @code{4*7-2} or @code{SIN(3.14)}.  A non-constant argument is an
expression that refers to values that may change from row to row, such as
column names or functions that are called with non-constant arguments.

For each invocation of the main function, @code{args->args} contains the
actual arguments that are passed for the row currently being processed.

Functions can refer to an argument @code{i} as follows:

@itemize @bullet
@item
An argument of type @code{STRING_RESULT} is given as a string pointer plus a
length, to allow handling of binary data or data of arbitrary length.  The
string contents are available as @code{args->args[i]} and the string length
is @code{args->lengths[i]}.  You should not assume that strings are
null-terminated.

@item
For an argument of type @code{INT_RESULT}, you must cast
@code{args->args[i]} to a @code{long long} value:

@example
long long int_val;
int_val = *((long long*) args->args[i]);
@end example

@item
For an argument of type @code{REAL_RESULT}, you must cast
@code{args->args[i]} to a @code{double} value:

@example
double    real_val;
real_val = *((double*) args->args[i]);
@end example
@end itemize

@item unsigned long *lengths
For the initialisation function, the @code{lengths} array indicates the
maximum string length for each argument. You should not change these.
For each invocation of the main function, @code{lengths} contains the
actual lengths of any string arguments that are passed for the row
currently being processed.  For arguments of types @code{INT_RESULT} or
@code{REAL_RESULT}, @code{lengths} still contains the maximum length of
the argument (as for the initialisation function).
@end table


@node UDF return values, UDF compiling, UDF arguments, Adding UDF
@subsubsection Return Values and Error Handling

@cindex UDFs, return values
@cindex return values, UDFs
@cindex errors, handling for UDFs
@cindex handling, errors

The initialisation function should return @code{0} if no error occurred and
@code{1} otherwise.  If an error occurs, @code{xxx_init()} should store a
null-terminated error message in the @code{message} parameter.  The message
will be returned to the client.  The message buffer is
@code{MYSQL_ERRMSG_SIZE} characters long, but you should try to keep the
message to less than 80 characters so that it fits the width of a standard
terminal screen.

The return value of the main function @code{xxx()} is the function value, for
@code{long long} and @code{double} functions.  A string functions should
return a pointer to the result and store the length of the string in the
@code{length} arguments.

Set these to the contents and length of the return value.  For example:

@example
memcpy(result, "result string", 13);
*length = 13;
@end example

The @code{result} buffer that is passed to the calc function is 255 byte
big. If your result fits in this, you don't have to worry about memory
allocation for results.

If your string function needs to return a string longer than 255 bytes,
you must allocate the space for it with @code{malloc()} in your
@code{xxx_init()} function or your @code{xxx()} function and free it in
your @code{xxx_deinit()} function.  You can store the allocated memory
in the @code{ptr} slot in the @code{UDF_INIT} structure for reuse by
future @code{xxx()} calls.  @xref{UDF calling}.

To indicate a return value of @code{NULL} in the main function, set
@code{is_null} to @code{1}:

@example
*is_null = 1;
@end example

To indicate an error return in the main function, set the @code{error}
parameter to @code{1}:

@example
*error = 1;
@end example

If @code{xxx()} sets @code{*error} to @code{1} for any row, the function
value is @code{NULL} for the current row and for any subsequent rows
processed by the statement in which @code{XXX()} was invoked.  (@code{xxx()}
will not even be called for subsequent rows.) @strong{Note}: in
MySQL versions prior to 3.22.10, you should set both @code{*error}
and @code{*is_null}:

@example
*error = 1;
*is_null = 1;
@end example


@node UDF compiling,  , UDF return values, Adding UDF
@subsubsection Compiling and Installing User-definable Functions

@cindex compiling, user-defined functions
@cindex UDFs, compiling
@cindex installing, user-defined functions

Files implementing UDFs must be compiled and installed on the host where the
server runs.  This process is described below for the example UDF file
@file{udf_example.cc} that is included in the MySQL source
distribution.  This file contains the following functions:

@itemize @bullet
@item
@code{metaphon()} returns a metaphon string of the string argument.
This is something like a soundex string, but it's more tuned for English.
@item
@code{myfunc_double()} returns the sum of the ASCII values of the
characters in its arguments, divided by the sum of the length of its arguments.
@item
@code{myfunc_int()} returns the sum of the length of its arguments.
@item
@code{sequence([const int])} returns an sequence starting from the given
number or 1 if no number has been given.
@item
@code{lookup()} returns the IP number for a hostname.
@item
@code{reverse_lookup()} returns the hostname for an IP number.
The function may be called with a string @code{"xxx.xxx.xxx.xxx"} or
four numbers.
@end itemize

A dynamically loadable file should be compiled as a sharable object file,
using a command something like this:

@example
shell> gcc -shared -o udf_example.so myfunc.cc
@end example

You can easily find out the correct compiler options for your system by
running this command in the @file{sql} directory of your MySQL
source tree:

@example
shell> make udf_example.o
@end example

You should run a compile command similar to the one that @code{make} displays,
except that you should remove the @code{-c} option near the end of the line
and add @code{-o udf_example.so} to the end of the line.  (On some systems,
you may need to leave  the @code{-c} on the command.)

Once you compile a shared object containing UDFs, you must install it
and tell MySQL about it.  Compiling a shared object from
@file{udf_example.cc} produces a file named something like
@file{udf_example.so} (the exact name may vary from platform to platform).
Copy this file to some directory searched by @code{ld}, such as
@file{/usr/lib}.  On many systems, you can set the @code{LD_LIBRARY} or
@code{LD_LIBRARY_PATH} environment variable to point at the directory where
you have your UDF function files. The @code{dlopen} manual page tells you
which variable you should use on your system. You should set this in
@code{mysql.server} or @code{safe_mysqld} and restart @code{mysqld}.

After the library is installed, notify @code{mysqld} about the new
functions with these commands:

@example
mysql> CREATE FUNCTION metaphon RETURNS STRING SONAME "udf_example.so";
mysql> CREATE FUNCTION myfunc_double RETURNS REAL SONAME "udf_example.so";
mysql> CREATE FUNCTION myfunc_int RETURNS INTEGER SONAME "udf_example.so";
mysql> CREATE FUNCTION lookup RETURNS STRING SONAME "udf_example.so";
mysql> CREATE FUNCTION reverse_lookup
    ->        RETURNS STRING SONAME "udf_example.so";
mysql> CREATE AGGREGATE FUNCTION avgcost
    ->        RETURNS REAL SONAME "udf_example.so";
@end example

Functions can be deleted using @code{DROP FUNCTION}:

@example
mysql> DROP FUNCTION metaphon;
mysql> DROP FUNCTION myfunc_double;
mysql> DROP FUNCTION myfunc_int;
mysql> DROP FUNCTION lookup;
mysql> DROP FUNCTION reverse_lookup;
mysql> DROP FUNCTION avgcost;
@end example

The @code{CREATE FUNCTION} and @code{DROP FUNCTION} statements update the
system table @code{func} in the @code{mysql} database.  The function's name,
type and shared library name are saved in the table.  You must have the
@strong{insert} and @strong{delete} privileges for the @code{mysql} database
to create and drop functions.

You should not use @code{CREATE FUNCTION} to add a function that has already
been created.  If you need to reinstall a function, you should remove it with
@code{DROP FUNCTION} and then reinstall it with @code{CREATE FUNCTION}.  You
would need to do this, for example, if you recompile a new version of your
function, so that @code{mysqld} gets the new version.  Otherwise, the server
will continue to use the old version.

Active functions are reloaded each time the server starts, unless you start
@code{mysqld} with the @code{--skip-grant-tables} option.  In this case, UDF
initialisation is skipped and UDFs are unavailable.  (An active function is
one that has been loaded with @code{CREATE FUNCTION} and not removed with
@code{DROP FUNCTION}.)


@node Adding native function,  , Adding UDF, Adding functions
@subsection Adding a New Native Function

@cindex adding, native functions
@cindex native functions, adding
@cindex functions, native, adding

The procedure for adding a new native function is described here.  Note
that you cannot add native functions to a binary distribution because
the procedure involves modifying MySQL source code.  You must
compile MySQL yourself from a source distribution.  Also note
that if you migrate to another version of MySQL (for example,
when a new version is released), you will need to repeat the procedure
with the new version.

To add a new native MySQL function, follow these steps:

@enumerate
@item
Add one line to @file{lex.h} that defines the function name in the
@code{sql_functions[]} array.
@item
If the function prototype is simple (just takes zero, one, two or three
arguments), you should in lex.h specify SYM(FUNC_ARG#) (where # is the
number of arguments) as the second argument in the
@code{sql_functions[]} array and add a function that creates a function
object in @file{item_create.cc}.  Take a look at @code{"ABS"} and
@code{create_funcs_abs()} for an example of this.

If the function prototype is complicated (for example takes a variable number
of arguments), you should add two lines to @file{sql_yacc.yy}. One
indicates the preprocessor symbol that @code{yacc} should define (this
should be added at the beginning of the file). Then define the function
parameters and add an ``item'' with these parameters to the
@code{simple_expr} parsing rule.  For an example, check all occurrences
of @code{ATAN} in @file{sql_yacc.yy} to see how this is done.
@item
In @file{item_func.h}, declare a class inheriting from @code{Item_num_func} or
@code{Item_str_func}, depending on whether your function returns a number or a
string.
@item
In @file{item_func.cc}, add one of the following declarations, depending
on whether you are defining a numeric or string function:
@example
double   Item_func_newname::val()
longlong Item_func_newname::val_int()
String  *Item_func_newname::Str(String *str)
@end example

If you inherit your object from any of the standard items (like
@code{Item_num_func} you probably only have to define one of the above
functions and let the parent object take care of the other functions.
For example, the @code{Item_str_func} class defines a @code{val()} function
that executes @code{atof()} on the value returned by @code{::str()}.

@item
You should probably also define the following object function:
@example
void Item_func_newname::fix_length_and_dec()
@end example
This function should at least calculate @code{max_length} based on the
given arguments. @code{max_length} is the maximum number of characters
the function may return.  This function should also set @code{maybe_null
= 0} if the main function can't return a @code{NULL} value.  The
function can check if any of the function arguments can return
@code{NULL} by checking the arguments @code{maybe_null} variable. You
can take a look at @code{Item_func_mod::fix_length_and_dec} for a
typical example of how to do this.
@end enumerate

All functions must be thread-safe (in other words, don't use any global or
static variables in the functions without protecting them with mutexes).

If you want to return @code{NULL}, from @code{::val()}, @code{::val_int()}
or @code{::str()} you should set @code{null_value} to 1 and return 0.

For @code{::str()} object functions, there are some additional
considerations to be aware of:

@itemize @bullet
@item
The @code{String *str} argument provides a string buffer that may be
used to hold the result. (For more information about the @code{String} type,
take a look at the @file{sql_string.h} file.)
@item
The @code{::str()} function should return the string that holds the result or
@code{(char*) 0} if the result is @code{NULL}.
@item
All current string functions try to avoid allocating any memory unless
absolutely necessary!
@end itemize


@node Adding procedures,  , Adding functions, Extending MySQL
@section Adding New Procedures to MySQL

@cindex procedures, adding
@cindex adding, procedures
@cindex new procedures, adding

In MySQL, you can define a procedure in C++ that can access and
modify the data in a query before it is sent to the client.  The modification
can be done on row-by-row or @code{GROUP BY} level.

We have created an example procedure in MySQL Version 3.23 to
show you what can be done.

Additionally we recommend you to take a look at @code{mylua}.
With this you can use the LUA language to load a procedure at
runtime into @code{mysqld}.

@menu
* procedure analyse::           Procedure analyse
* Writing a procedure::         Writing a procedure.
@end menu


@node procedure analyse, Writing a procedure, Adding procedures, Adding procedures
@subsection Procedure Analyse

@code{analyse([max elements,[max memory]])}

This procedure is defined in the @file{sql/sql_analyse.cc}.  This
examines the result from your query and returns an analysis of the
results:

@itemize @bullet
@item
@code{max elements} (default 256) is the maximum number of distinct values
@code{analyse} will notice per column.  This is used by @code{analyse} to check if
the optimal column type should be of type @code{ENUM}.
@item
@code{max memory} (default 8192) is the maximum memory @code{analyse} should
allocate per column while trying to find all distinct values.
@end itemize

@example
SELECT ... FROM ... WHERE ... PROCEDURE ANALYSE([max elements,[max memory]])
@end example


@node Writing a procedure,  , procedure analyse, Adding procedures
@subsection Writing a Procedure

For the moment, the only documentation for this is the source.

You can find all information about procedures by examining the following files:

@itemize @bullet
@item @file{sql/sql_analyse.cc}
@item @file{sql/procedure.h}
@item @file{sql/procedure.cc}
@item @file{sql/sql_select.cc}
@end itemize


@node Problems, Contrib, Extending MySQL, Top
@appendix Problems and Common Errors

@cindex problems, common errors
@cindex errors, common

@menu
* What is crashing::            How to determine what is causing problems
* Common errors::               Common Errors When Using MySQL
* Installation Issues::         Installation Related Issues
* Administration Issues::       Administration Related Issues
* Query Issues::                Query Related Issues
* Table Definition Issues::     Table Definition Related Issues
@end menu

This chapter lists some common problems and error messages that users have
run into.  You will learn how to figure out what the problem is, and what
to do to solve it.  You will also find proper solutions to some common
problems.


@node What is crashing, Common errors, Problems, Problems
@appendixsec How to Determine What Is Causing Problems

When you run into problems, the first thing you should do is to find out
which program / piece of equipment is causing problems:

@itemize @bullet
@item
If you have one of the following symptoms, then it is probably a hardware
(like memory, motherboard, CPU, or hard disk) or kernel problem:
@itemize @minus
@item
The keyboard doesn't work.  This can normally be checked by pressing
Caps Lock. If the Caps Lock light doesn't change you have to replace
your keyboard.  (Before doing this, you should try to reboot
your computer and check all cables to the keyboard.)
@item
The mouse pointer doesn't move.
@item
The machine doesn't answer to a remote machine's pings.
@item
Different, unrelated programs don't behave correctly.
@item
If your system rebooted unexpectedly (a faulty user level program should
@strong{never} be able to take down your system).
@end itemize

In this case you should start by checking all your cables and run some
diagnostic tool to check your hardware!
You should also check if there are any patches, updates, or service
packs for your operating system that could likely solve your problems.
Check also that all your libraries (like glibc) are up to date.

It's always good to use a machine with ECC memory to discover
memory problems early!
@item
If your keyboard is locked up, you may be able to fix this by
logging into your machine from another machine and execute
@code{kbd_mode -a} on it.

@item
Please examine your system log file (/var/log/messages or similar) for
reasons for your problems.  If you think the problem is in MySQL
then you should also examine MySQL's log files. @xref{Update log}.

@item
If you don't think you have hardware problems, you should try to find
out which program is causing problems.

Try using @code{top}, @code{ps}, @code{taskmanager}, or some similar program,
to check which program is taking all CPU or is locking the machine.

@item
Check with @code{top}, @code{df}, or a similar program if you are out of
memory, disk space, open files, or some other critical resource.

@item
If the problem is some runaway process, you can always try to kill it. If it
doesn't want to die, there is probably a bug in the operating system.
@end itemize

If after you have examined all other possibilities and you have
concluded that it's the MySQL server or a MySQL client
that is causing the problem, it's time to do a bug report for our
mailing list or our support team.  In the bug report, try to give a
very detailed description of how the system is behaving and what you think is
happening. You should also state why you think it's MySQL that
is causing the problems.  Take into consideration all the situations in
this chapter. State any problems exactly how they appear when you
examine your system. Use the 'cut and paste' method for any output
and/or error messages from programs and/or log files!

Try to describe in detail which program is not working and all
symptoms you see!  We have in the past received many bug reports that just
state "the system doesn't work".  This doesn't provide us with any
information about what could be the problem.

If a program fails, it's always useful to know:

@itemize @bullet
@item
Has the program in question made a segmentation fault (core dumped)?
@item
Is the program taking up the whole CPU? Check with @code{top}. Let the
program run for a while, it may be evaluating something heavy.
@item
If it's the @code{mysqld} server that is causing problems, can you
do @code{mysqladmin -u root ping} or @code{mysqladmin -u root processlist}?
@item
What does a client program say (try with @code{mysql}, for example)
when you try to connect to the MySQL server?
Does the client jam?  Do you get any output from the program?
@end itemize

When sending a bug report, you should of follow the outlines
described in this manual. @xref{Asking questions}.


@node Common errors, Installation Issues, What is crashing, Problems
@appendixsec Common Errors When Using MySQL

@cindex errors, list of

@menu
* Error Access denied::         @code{Access denied} Error
* Gone away::                   @code{MySQL server has gone away} error
* Can not connect to server::   @code{Can't connect to [local] MySQL server} error
* Blocked host::                @code{Host '...' is blocked} error
* Too many connections::        @code{Too many connections} error
* Non-transactional tables::    @code{Some non-transactional changed tables couldn't be rolled back} Error
* Out of memory::               @code{Out of memory} error
* Packet too large::            @code{Packet too large} error
* Communication errors::        Communication errors / Aborted connection
* Full table::                  @code{The table is full} error
* Cannot create::               @code{Can't create/write to file} Error
* Commands out of sync::        @code{Commands out of sync} error in client
* Ignoring user::               @code{Ignoring user} error
* Cannot find table::           @code{Table 'xxx' doesn't exist} error
* Cannot initialize character set::  @code{Can't initialize character set xxx} error
* Not enough file handles::     File Not Found
@end menu


This section lists some errors that users frequently get.  You will find
descriptions of the errors, and how to solve the problem here.


@node Error Access denied, Gone away, Common errors, Common errors
@appendixsubsec @code{Access denied} Error

@cindex errors, access denied
@cindex problems, access denied errors
@cindex access denied errors

@xref{Privileges}, and especially. @xref{Access denied}.


@node Gone away, Can not connect to server, Error Access denied, Common errors
@appendixsubsec @code{MySQL server has gone away} Error

This section also covers the related @code{Lost connection to server
during query} error.

The most common reason for the @code{MySQL server has gone away} error
is that the server timed out and closed the connection. By default, the
server closes the connection after 8 hours if nothing has happened. You
can change the time limit by setting the @code{wait_timeout} variable when
you start @code{mysqld}.

Another common reason to receive the @code{MySQL server has gone away} error
is because you have issued a ``close'' on your MySQL connection
and then tried to run a query on the closed connection.

If you have a script, you just have to issue the query again for the client
to do an automatic reconnection.

You normally can get the following error codes in this case
(which one you get is OS-dependent):

@multitable @columnfractions .35 .65
@item @strong{Error code} @tab @strong{Description}
@item @code{CR_SERVER_GONE_ERROR} @tab The client couldn't send a question to the
server.
@item @code{CR_SERVER_LOST} @tab  The client didn't get an error when writing
to the server, but it didn't get a full answer (or any answer) to the question.
@end multitable

You will also get this error if someone has kills the running thread with
@code{kill #threadid#}.

You can check that the MySQL hasn't died by executing @code{mysqladmin
version} and examining the uptime.  If the problem is that mysqld
crashed you should concentrate one finding the reason for the crash.
You should in this case start by checking if issuing the query again
will kill MySQL again. @xref{Crashing}.

You can also get these errors if you send a query to the server that is
incorrect or too large. If @code{mysqld} gets a packet that is too large
or out of order, it assumes that something has gone wrong with the client and
closes the connection.  If you need big queries (for example, if you are
working with big @code{BLOB} columns), you can increase the query limit by
starting @code{mysqld} with the @code{-O max_allowed_packet=#} option
(default 1M). The extra memory is allocated on demand, so @code{mysqld} will
use more memory only when you issue a big query or when @code{mysqld} must
return a big result row!

If you want to make a bug report regarding this problem, be sure that
you include the following information:

@itemize @bullet
@item
Include information if MySQL died or not. (You can find this in the
@code{hostname.err file}. @xref{Crashing}.
@item
If a specific query kills @code{mysqld} and the involved tables where
checked with @code{CHECK TABLE} before you did the query, can you do
a test case for this? @xref{Reproduceable test case}.
@item
What is the value of the @code{wait_timeout} variable in the MySQL server ?
@code{mysqladmin variables} gives you the value of this
@item
Have you tried to run @code{mysqld} with @code{--log} and check if the
issued query appears in the log ?
@end itemize

@xref{Asking questions}.

@node Can not connect to server, Blocked host, Gone away, Common errors
@appendixsubsec @code{Can't connect to [local] MySQL server} Error

A MySQL client on Unix can connect to the @code{mysqld} server in two
different ways: Unix sockets, which connect through a file in the file
system (default @file{/tmp/mysqld.sock}) or TCP/IP, which connects
through a port number.  Unix sockets are faster than TCP/IP but can only
be used when connecting to a server on the same computer.  Unix sockets
are used if you don't specify a hostname or if you specify the special
hostname @code{localhost}.

On Windows, if the @code{mysqld} server is running on 9x/Me, you can
connect only via TCP/IP. If the server is running on NT/2000/XP and
mysqld is started with @code{--enable-named-pipe}, you
can also connect with named pipes. The name of the named pipe is MySQL.
If you don't give a hostname when connecting to @code{mysqld}, a MySQL
client will first try to connect to the named pipe, and if this doesn't
work it will connect to the TCP/IP port.  You can force the use of named
pipes on Windows by using @code{.} as the hostname.

The error (2002) @code{Can't connect to ...} normally means that there
isn't a MySQL server running on the system or that you are
using a wrong socket file or TCP/IP port when trying to connect to the
@code{mysqld} server.

Start by checking (using @code{ps} or the task manager on Windows) that
there is a process running named @code{mysqld} on your server!  If there
isn't any @code{mysqld} process, you should start one. @xref{Starting
server}.

If a @code{mysqld} process is running, you can check the server by
trying these different connections (the port number and socket pathname
might be different in your setup, of course):

@example
shell> mysqladmin version
shell> mysqladmin variables
shell> mysqladmin -h `hostname` version variables
shell> mysqladmin -h `hostname` --port=3306 version
shell> mysqladmin -h 'ip for your host' version
shell> mysqladmin --socket=/tmp/mysql.sock version
@end example

Note the use of backquotes rather than forward quotes with the @code{hostname}
command; these cause the output of @code{hostname} (that is, the current
hostname) to be substituted into the @code{mysqladmin} command.

Here are some reasons the @code{Can't connect to local MySQL server}
error might occur:

@itemize @bullet
@item
@code{mysqld} is not running.
@item
You are running on a system that uses MIT-pthreads.
If you are running on a system that doesn't have native threads,
@code{mysqld} uses the MIT-pthreads package.  @xref{Which OS}.  However,
all MIT-pthreads versions doesn't support Unix sockets. On a system
without sockets support you must always specify the hostname explicitly
when connecting to the server. Try using this command to check the
connection to the server:
@example
shell> mysqladmin -h `hostname` version
@end example
@item
Someone has removed the Unix socket that @code{mysqld} uses (default
@file{/tmp/mysqld.sock}).  You might have a @code{cron} job that removes
the MySQL socket (for example, a job that removes old files
from the @file{/tmp} directory).  You can always run @code{mysqladmin
version} and check that the socket @code{mysqladmin} is trying to use
really exists.  The fix in this case is to change the @code{cron} job to
not remove @file{mysqld.sock} or to place the socket somewhere else.
@xref{Problems with mysql.sock}.
@item
You have started the @code{mysqld} server with
the @code{--socket=/path/to/socket} option.  If you change the socket
pathname for the server, you must also notify the MySQL clients
about the new path. You can do this by providing the socket path
as an argument to the client. @xref{Problems with mysql.sock}.
@item
You are using Linux and one thread has died (core dumped).  In this case
you must kill the other @code{mysqld} threads (for example, with the
@code{mysql_zap} script before you can start a new MySQL
server.  @xref{Crashing}.
@item
You may not have read and write privilege to either the directory that holds
the socket file or privilege to the socket file itself. In this case you
have to either change the privilege for the directory / file or restart
@code{mysqld} so that it uses a directory that you can access.
@end itemize

If you get the error message @code{Can't connect to MySQL server on
some_hostname}, you can try the following things to find out what the
problem is :

@itemize @bullet
@item
Check if the server is up by doing @code{telnet your-host-name
tcp-ip-port-number} and press Enter a couple of times.  If there
is a MySQL server running on this port you should get a
responses that includes the version number of the running MySQL
server.  If you get an error like @code{telnet: Unable to connect to
remote host: Connection refused}, then there is no server running on the
given port.
@item
Try connecting to the @code{mysqld} daemon on the local machine and check
the TCP/IP port that @code{mysqld} it's configured to use (variable @code{port}) with
@code{mysqladmin variables}.
@item
Check that your @code{mysqld} server is not started with the
@code{--skip-networking} option.
@end itemize


@node Blocked host, Too many connections, Can not connect to server, Common errors
@appendixsubsec @code{Host '...' is blocked} Error

If you get an error like this:

@example
Host 'hostname' is blocked because of many connection errors.
Unblock with 'mysqladmin flush-hosts'
@end example

this means that @code{mysqld} has gotten a lot (@code{max_connect_errors})
of connect requests from the host @code{'hostname'} that have been interrupted
in the middle. After @code{max_connect_errors} failed requests, @code{mysqld}
assumes that something is wrong (like an attack from a cracker), and
blocks the site from further connections until someone executes the command
@code{mysqladmin flush-hosts}.

By default, @code{mysqld} blocks a host after 10 connection errors.
You can easily adjust this by starting the server like this:

@example
shell> safe_mysqld -O max_connect_errors=10000 &
@end example

Note that if you get this error message for a given host, you should first
check that there isn't anything wrong with TCP/IP connections from that
host.  If your TCP/IP connections aren't working, it won't do you any good to
increase the value of the @code{max_connect_errors} variable!


@node Too many connections, Non-transactional tables, Blocked host, Common errors
@appendixsubsec @code{Too many connections} Error

If you get the error @code{Too many connections} when you try to connect
to MySQL, this means that there is already @code{max_connections}
clients connected to the @code{mysqld} server.

If you need more connections than the default (100), then you should restart
@code{mysqld} with a bigger value for the @code{max_connections} variable.

Note that @code{mysqld} actually allows (@code{max_connections}+1)
clients to connect.  The last connection is reserved for a user with the
@strong{process} privilege.  By not giving this privilege to normal
users (they shouldn't need this), an administrator with this privilege
can log in and use @code{SHOW PROCESSLIST} to find out what could be
wrong. @xref{SHOW}.

The maximum number of connects MySQL is depending on how good
the thread library is on a given platform. Linux or Solaris should be
able to support 500-1000 simultaneous connections, depending on how much
RAM you have and what your clients are doing.


@node Non-transactional tables, Out of memory, Too many connections, Common errors
@appendixsubsec @code{Some non-transactional changed tables couldn't be rolled back} Error

@cindex Non-transactional tables

If you get the error/warning: @code{Warning: Some non-transactional
changed tables couldn't be rolled back} when trying to do a
@code{ROLLBACK}, this means that some of the tables you used in the
transaction didn't support transactions.  These non-transactional tables
will not be affected by the @code{ROLLBACK} statement.

The most typical case when this happens is when you have tried to create
a table of a type that is not supported by your @code{mysqld} binary.
If @code{mysqld} doesn't support a table type (or if the table type is
disabled by a startup option) , it will instead create the table type
with the table type that is most resembles to the one you requested,
probably @code{MyISAM}.

You can check the table type for a table by doing:

@code{SHOW TABLE STATUS LIKE 'table_name'}. @xref{SHOW TABLE STATUS}.

You can check the extensions your @code{mysqld} binary supports by doing:

@code{show variables like 'have_%'}. @xref{SHOW VARIABLES}.


@node Out of memory, Packet too large, Non-transactional tables, Common errors
@appendixsubsec @code{Out of memory} Error

If you issue a query and get something like the following error:

@example
mysql: Out of memory at line 42, 'malloc.c'
mysql: needed 8136 byte (8k), memory in use: 12481367 bytes (12189k)
ERROR 2008: MySQL client ran out of memory
@end example

note that the error refers to the MySQL client @code{mysql}. The
reason for this error is simply that the client does not have enough memory to
store the whole result.

To remedy the problem, first check that your query is correct. Is it
reasonable that it should return so many rows?  If so,
you can use @code{mysql --quick}, which uses @code{mysql_use_result()}
to retrieve the result set.  This places less of a load on the client (but
more on the server).


@node Packet too large, Communication errors, Out of memory, Common errors
@appendixsubsec @code{Packet too large} Error

When a MySQL client or the @code{mysqld} server gets a packet bigger
than @code{max_allowed_packet} bytes, it issues a @code{Packet too large}
error and closes the connection.

In MySQL 3.23 the biggest possible packet is 16M (due to limits in the
client/server protocol). In MySQL 4.0.1 and up, this is only limited by
the amount on memory you have on your server (up to a theoretical
maximum of 2G).

A communication packet is a single SQL statement sent to the MySQL server
or a single row that is sent to the client.

When a MySQL client or the @code{mysqld} server gets a packet bigger
than @code{max_allowed_packet} bytes, it issues a @code{Packet too
large} error and closes the connection.  With some clients, you may also
get @code{Lost connection to MySQL server during query} error if the
communication packet is too big.

Note that both the client and the server has it's own
@code{max_allowed_packet} variable. If you want to handle big packets,
you have to increase this variable both in the client and in the server.

It's safe to increase this variable as memory is only allocated when
needed; this variable is more a precaution to catch wrong packets
between the client/server and also to ensure that you don't accidently
use big packets so that you run out of memory.

If you are using the @code{mysql} client, you may specify a bigger
buffer by starting the client with @code{mysql --set-variable=max_allowed_packet=8M}.  Other clients have different methods to set this variable.

You can use the option file to set @code{max_allowed_packet} to a larger
size in @code{mysqld}.  For example, if you are expecting to store the
full length of a @code{MEDIUMBLOB} into a table, you'll need to start
the server with the @code{set-variable=max_allowed_packet=16M} option.

You can also get strange problems with large packets if you are using
big blobs, but you haven't given @code{mysqld} access to enough memory
to handle the query.  If you suspect this is the case, try adding
@code{ulimit -d 256000} to the beginning of the @code{safe_mysqld} script
and restart @code{mysqld}.

@node Communication errors, Full table, Packet too large, Common errors
@appendixsubsec Communication Errors / Aborted Connection

@cindex aborted clients
@cindex aborted connection
@cindex connection, aborted

Starting with @code{MySQL 3.23.40} you only get the @code{Aborted
connection} error of you start @code{mysqld} with @code{--warnings}.

If you find errors like the following in your error log.

@example
010301 14:38:23  Aborted connection 854 to db: 'users' user: 'josh'
@end example

@xref{Error log}.

This means that something of the following has happened:

@itemize @bullet
@item
The client program did not call @code{mysql_close()} before exit.
@item
The client had been sleeping more than @code{wait_timeout} or
@code{interactive_timeout} without doing any requests. @xref{SHOW
VARIABLES}.
@item
The client program ended abruptly in the middle of the transfer.
@end itemize

When the above happens, the server variable @code{Aborted_clients} is
incremented.

The server variable @code{Aborted_connects} is incremented when:

@itemize @bullet
@item
When a connection packet doesn't contain the right information.
@item
When the user didn't have privileges to connect to a database.
@item
When a user uses a wrong password.
@item
When it takes more than @code{connect_timeout} seconds to get
a connect package.
@end itemize

Note that the above could indicate that someone is trying to break into
your database!

@xref{SHOW VARIABLES}.

Other reasons for problems with Aborted clients / Aborted connections.
@itemize @bullet
@item
Usage of duplex Ethernet protocol, both half and full with
Linux. Many Linux Ethernet drivers have this bug. You should test
for this bug by transferring a huge file via ftp between these two
machines. If a transfer goes in burst-pause-burst-pause ... mode then
you are experiencing a Linux duplex syndrome. The only solution to
this problem is switching of both half and full duplexing on hubs
and switches.
@item
Some problem with the thread library that causes interrupts on reads.
@item
Badly configured TCP/IP.
@item
Faulty Ethernets or hubs or switches, cables ... This can be diagnosed
properly only by replacing hardware.
@item
@code{max_allowed_packet} is too small or queries require more memory
than you have alloacated for @code{mysqld}. @xref{Packet too large}.
@end itemize


@node Full table, Cannot create, Communication errors, Common errors
@appendixsubsec @code{The table is full} Error

@cindex table is full

This error occurs in older MySQL versions when an in-memory temporary
table becomes larger than @code{tmp_table_size} bytes. To avoid this
problem, you can use the @code{-O tmp_table_size=#} option to
@code{mysqld} to increase the temporary table size or use the SQL
option @code{SQL_BIG_TABLES} before you issue the problematic
query. @xref{SET OPTION, , @code{SET OPTION}}.

You can also start @code{mysqld} with the @code{--big-tables} option.
This is exactly the same as using @code{SQL_BIG_TABLES} for all queries.

In MySQL Version 3.23, in-memory temporary tables will automatically be
converted to a disk-based @code{MyISAM} table after the table size gets
bigger than @code{tmp_table_size}.


@node Cannot create, Commands out of sync, Full table, Common errors
@appendixsubsec @code{Can't create/write to file} Error

@cindex can't create/write to file

If you get an error for some queries of type:

@example
Can't create/write to file '\\sqla3fe_0.ism'.
@end example

this means that MySQL can't create a temporary file for the
result set in the given temporary directory. (The above error is a
typical error message on Windows, and the Unix error message is similar.)
The fix is to start @code{mysqld} with @code{--tmpdir=path} or to add to your option
file:

@example
[mysqld]
tmpdir=C:/temp
@end example

assuming that the @file{c:\\temp} directory exists. @xref{Option files}.

Check also the error code that you get with @code{perror}. One reason
may also be a disk full error;

@example
shell> perror 28
Error code  28:  No space left on device
@end example


@node Commands out of sync, Ignoring user, Cannot create, Common errors
@appendixsubsec @code{Commands out of sync} Error in Client

@cindex commands out of sync

If you get @code{Commands out of sync; you can't run this command now}
in your client code, you are calling client functions in the wrong order!

This can happen, for example, if you are using @code{mysql_use_result()} and
try to execute a new query before you have called @code{mysql_free_result()}.
It can also happen if you try to execute two queries that return data without
a @code{mysql_use_result()} or @code{mysql_store_result()} in between.


@node Ignoring user, Cannot find table, Commands out of sync, Common errors
@appendixsubsec @code{Ignoring user} Error

If you get the following error:

@code{Found wrong password for user: 'some_user@@some_host'; ignoring user}

this means that when @code{mysqld} was started or when it reloaded the
permissions tables, it found an entry in the @code{user} table with
an invalid password.  As a result, the entry is simply ignored by the
permission system.

Possible causes of and fixes for this problem:

@itemize @bullet
@item
You may be running a new version of @code{mysqld} with an old
@code{user} table.
You can check this by executing @code{mysqlshow mysql user} to see if
the password field is shorter than 16 characters. If so, you can correct this
condition by running the @code{scripts/add_long_password} script.

@item
The user has an old password (8 characters long) and you didn't start
@code{mysqld} with the @code{--old-protocol} option.
Update the user in the @code{user} table with a new password or
restart @code{mysqld} with @code{--old-protocol}.

@item
@findex PASSWORD()
You have specified a password in the @code{user} table without using the
@code{PASSWORD()} function.  Use @code{mysql} to update the user in the
@code{user} table with a new password. Make sure to use the @code{PASSWORD()}
function:

@example
mysql> UPDATE user SET password=PASSWORD('your password')
    ->             WHERE user='XXX';
@end example
@end itemize


@node Cannot find table, Cannot initialize character set, Ignoring user, Common errors
@appendixsubsec @code{Table 'xxx' doesn't exist} Error

If you get the error @code{Table 'xxx' doesn't exist} or @code{Can't
find file: 'xxx' (errno: 2)}, this means that no table exists
in the current database with the name @code{xxx}.

Note that as MySQL uses directories and files to store databases and
tables, the database and table names are @strong{case-sensitive}!
(On Windows the databases and tables names are not case-sensitive, but all
references to a given table within a query must use the same case!)

You can check which tables you have in the current database with
@code{SHOW TABLES}. @xref{SHOW, , @code{SHOW}}.


@node Cannot initialize character set, Not enough file handles, Cannot find table, Common errors
@appendixsubsec @code{Can't initialize character set xxx} error

@cindex multibyte character sets

If you get an error like:

@example
MySQL Connection Failed: Can't initialize character set xxx
@end example

This means one of the following things:

@itemize @bullet
@item
The character set is a multi-byte character set and you have no support
for the character set in the client.

In this case you need to recompile the client with
@code{--with-charset=xxx} or with @code{--with-extra-charsets=xxx}.
@xref{configure options}.

All standard MySQL binaries are compiled with
@code{--with-extra-character-sets=complex} which will enable support for
all multi-byte character sets. @xref{Character sets}.

@item
The character set is a simple character set which is not compiled into
@code{mysqld} and the character set definition files are not in the place
where the client expects to find them.

In this case you need to:

@itemize @bullet
@item
Recompile the client with support for the character set.
@xref{configure options}.
@item
Specify to the client where the character set definition files are. For many
clients you can do this with the
@code{--character-sets-dir=path-to-charset-dir} option.
@item
Copy the character definition files to the path where the client expects them
to be.
@end itemize
@end itemize


@node Not enough file handles,  , Cannot initialize character set, Common errors
@appendixsubsec File Not Found

If you get @code{ERROR '...' not found (errno: 23)}, @code{Can't open
file: ... (errno: 24)}, or any other error with @code{errno 23} or
@code{errno 24} from MySQL, it means that you haven't allocated
enough file descriptors for MySQL.  You can use the
@code{perror} utility to get a description of what the error number
means:

@example
shell> perror 23
File table overflow
shell> perror 24
Too many open files
shell> perror 11
Resource temporarily unavailable
@end example

The problem here is that @code{mysqld} is trying to keep open too many
files simultaneously.  You can either tell @code{mysqld} not to open so
many files at once or increase the number of file descriptors
available to @code{mysqld}.

To tell @code{mysqld} to keep open fewer files at a time, you can make
the table cache smaller by using the @code{-O table_cache=32} option to
@code{safe_mysqld} (the default value is 64). Reducing the value of
@code{max_connections} will also reduce the number of open files (the
default value is 90).

@tindex ulimit
To change the number of file descriptors available to @code{mysqld}, you
can use the option @code{--open-files-limit=#} to @code{safe_mysqld} or
@code{-O open-files-limit=#} to @code{mysqld}. @xref{SHOW VARIABLES}.
The easiest way to do that is to add the option to your option file.
@xref{Option files}.  If you have an old @code{mysqld} version that
doesn't support this, you can edit the @code{safe_mysqld} script.  There
is a commented-out line @code{ulimit -n 256} in the script.  You can
remove the @code{'#'} character to uncomment this line, and change the
number 256 to affect the number of file descriptors available to
@code{mysqld}.

@code{ulimit} (and @code{open-files-limit}) can increase the number of
file descriptors, but only up to the limit imposed by the operating
system. There is also a 'hard' limit that can only be overrided if you
start @code{safe_mysqld} or @code{mysqld} as root (just remember that
you need to also use the @code{--user=...} option in this case).  If you
need to increase the OS limit on the number of file descriptors
available to each process, consult the documentation for your operating
system.

Note that if you run the @code{tcsh} shell, @code{ulimit} will not work!
@code{tcsh} will also report incorrect values when you ask for the current
limits!  In this case you should start @code{safe_mysqld} with @code{sh}!


@node Installation Issues, Administration Issues, Common errors, Problems
@appendixsec Installation Related Issues


@menu
* Link errors::                 Problems When Linking with the MySQL Client Library
* Changing MySQL user::         How to Run MySQL As a Normal User
* File permissions ::           Problems with File Permissions
@end menu

@node Link errors, Changing MySQL user, Installation Issues, Installation Issues
@appendixsubsec Problems When Linking with the MySQL Client Library

@cindex linking, errors
@cindex errors, linking
@cindex problems, linking

If you are linking your program and you get errors for unreferenced
symbols that start with @code{mysql_}, like the following:

@example
/tmp/ccFKsdPa.o: In function `main':
/tmp/ccFKsdPa.o(.text+0xb): undefined reference to `mysql_init'
/tmp/ccFKsdPa.o(.text+0x31): undefined reference to `mysql_real_connect'
/tmp/ccFKsdPa.o(.text+0x57): undefined reference to `mysql_real_connect'
/tmp/ccFKsdPa.o(.text+0x69): undefined reference to `mysql_error'
/tmp/ccFKsdPa.o(.text+0x9a): undefined reference to `mysql_close'
@end example

you should be able to solve this by adding @code{-Lpath-to-the-mysql-library
-lmysqlclient} @strong{last} on your link line.

If you get @code{undefined reference} errors for the @code{uncompress}
or @code{compress} function, add @code{-lz} @strong{last} on your link
line and try again!

If you get @code{undefined reference} errors for functions that should
exist on your system, like @code{connect}, check the man page for the
function in question, for which libraries you should add to the link
line!

If you get @code{undefined reference} errors for functions that don't
exist on your system, like the following:

@example
mf_format.o(.text+0x201): undefined reference to `__lxstat'
@end example

it usually means that your library is compiled on a system that is not
100 % compatible with yours.  In this case you should download the
latest MySQL source distribution and compile this yourself.
@xref{Installing source}.

If you are trying to run a program and you then get errors for
unreferenced symbols that start with @code{mysql_} or that the
@code{mysqlclient} library can't be found, this means that your system
can't find the share @code{libmysqlclient.so} library.

The fix for this is to tell your system to search after shared
libraries where the library is located by one of the following methods:

@itemize @bullet
@item
Add the path to the directory where you have @code{libmysqlclient.so} the
@code{LD_LIBRARY_PATH} environment variable.
@item
Add the path to the directory where you have @code{libmysqlclient.so} the
@code{LD_LIBRARY} environment variable.
@item
Copy @code{libmysqlclient.so} to some place that is searched by your system,
like @file{/lib}, and update the shared library information by executing
@code{ldconfig}.
@end itemize

Another way to solve this problem is to link your program statically, with
@code{-static}, or by removing the dynamic MySQL libraries
before linking your code. In the second case you should be
sure that no other programs are using the dynamic libraries!


@node Changing MySQL user, File permissions , Link errors, Installation Issues
@appendixsubsec How to Run MySQL As a Normal User

@cindex starting, @code{mysqld}
@cindex @code{mysqld}, starting

The MySQL server @code{mysqld} can be started and run by any user.
In order to change @code{mysqld} to run as a Unix user @code{user_name}, you must
do the following:

@enumerate
@item
Stop the server if it's running (use @code{mysqladmin shutdown}).

@item
Change the database directories and files so that @code{user_name} has
privileges to read and write files in them (you may need to do this as
the Unix @code{root} user):

@example
shell> chown -R user_name /path/to/mysql/datadir
@end example

If directories or files within the MySQL data directory are
symlinks, you'll also need to follow those links and change the directories
and files they point to.  @code{chown -R} may not follow symlinks for
you.

@item
Start the server as user @code{user_name}, or, if you are using
MySQL Version 3.22 or later, start @code{mysqld} as the Unix @code{root}
user and use the @code{--user=user_name} option.  @code{mysqld} will switch
to run as the Unix user @code{user_name} before accepting any connections.

@item
To start the server as the given user name automatically at system
startup time, add a @code{user} line that specifies the user name to
the @code{[mysqld]} group of the @file{/etc/my.cnf} option file or the
@file{my.cnf} option file in the server's data directory. For example:

@example
[mysqld]
user=user_name
@end example
@end enumerate

At this point, your @code{mysqld} process should be running fine and dandy as
the Unix user @code{user_name}. One thing hasn't changed, though: the
contents of the permissions tables. By default (right after running the
permissions table install script @code{mysql_install_db}), the MySQL
user @code{root} is the only user with permission to access the @code{mysql}
database or to create or drop databases.  Unless you have changed those
permissions, they still hold. This shouldn't stop you from accessing
MySQL as the MySQL @code{root} user when you're logged in
as a Unix user other than @code{root}; just specify the @code{-u root} option
to the client program.

Note that accessing MySQL as @code{root}, by supplying @code{-u
root} on the command-line, has @strong{nothing} to do with MySQL running
as the Unix @code{root} user, or, indeed, as another Unix user. The access
permissions and user names of MySQL are completely separate from
Unix user names. The only connection with Unix user names is that if you
don't provide a @code{-u} option when you invoke a client program, the client
will try to connect using your Unix login name as your MySQL user
name.

If your Unix box itself isn't secured, you should probably at least put a
password on the MySQL @code{root} users in the access tables.
Otherwise, any user with an account on that machine can run @code{mysql -u
root db_name} and do whatever he likes.


@node File permissions ,  , Changing MySQL user, Installation Issues
@appendixsubsec Problems with File Permissions

@cindex files, permissions
@cindex error mesaages, can't find file
@cindex files, not found message

If you have problems with file permissions, for example, if @code{mysql}
issues the following error message when you create a table:

@example
ERROR: Can't find file: 'path/with/filename.frm' (Errcode: 13)
@end example

@tindex UMASK environment variable
@tindex Environment variable, UMASK
then the environment variable @code{UMASK} might be set incorrectly when
@code{mysqld} starts up.  The default umask value is @code{0660}. You can
change this behavior by starting @code{safe_mysqld} as follows:

@example
shell> UMASK=384  # = 600 in octal
shell> export UMASK
shell> /path/to/safe_mysqld &
@end example

@tindex UMASK_DIR environment variable
@tindex Environment variable, UMASK_DIR
By default MySQL will create database and @code{RAID}
directories with permission type 0700.  You can modify this behavior by
setting the @code{UMASK_DIR} variable. If you set this, new
directories are created with the combined @code{UMASK} and
@code{UMASK_DIR}. For example, if you want to give group access to
all new directories, you can do:

@example
shell> UMASK_DIR=504  # = 770 in octal
shell> export UMASK_DIR
shell> /path/to/safe_mysqld &
@end example

In MySQL Version 3.23.25 and above, MySQL assumes that the
value for @code{UMASK} and @code{UMASK_DIR} is in octal if it starts
with a zero.

@xref{Environment variables}.


@node Administration Issues, Query Issues, Installation Issues, Problems
@appendixsec Administration Related Issues


@menu
* Crashing::                    What To Do If MySQL Keeps Crashing
* Resetting permissions::       How to Reset a Forgotten Password
* Full disk::                   How MySQL Handles a Full Disk
* Temporary files::             Where MySQL Stores Temporary Files
* Problems with mysql.sock::    How to Protect @file{/tmp/mysql.sock} from Being Deleted
* Timezone problems::           Time Zone Problems
@end menu

@node Crashing, Resetting permissions, Administration Issues, Administration Issues
@appendixsubsec What To Do If MySQL Keeps Crashing

@cindex crash, repeated

All MySQL versions are tested on many platforms before they are
released.  This doesn't mean that there aren't any bugs in
MySQL, but it means if there are bugs, they are very few and can be
hard to find.  If you have a problem, it will always help if you try to
find out exactly what crashes your system, as you will have a much better
chance of getting this fixed quickly.

First, you should try to find out whether the problem is that the
@code{mysqld} daemon dies or whether your problem has to do with your
client.  You can check how long your @code{mysqld} server has been up by
executing @code{mysqladmin version}.  If @code{mysqld} has died, you may
find the reason for this in the file
@file{mysql-data-directory/`hostname`.err}. @xref{Error log}.

Many crashes of MySQL are caused by corrupted index / data
files.  MySQL will update the data on disk, with the
@code{write()} system call, after every SQL statement and before the
client is notified about the result. (This is not true if you are running
with @code{delay_key_write}, in which case only the data is written.)
This means that the data is safe even if @code{mysqld} crashes, as the OS will
ensure that the not flushed data is written to disk.  You can force
MySQL to sync everything to disk after every SQL command by
starting @code{mysqld} with @code{--flush}.

The above means that normally you shouldn't get corrupted tables unless:

@itemize @bullet
@item
Someone/something killed @code{mysqld} or the machine in the middle
of an update.
@item
You have found a bug in @code{mysqld} that caused it to die in the
middle of an update.
@item
Someone is manipulating the data/index files outside of @strong{mysqld}
without locking the table properly.
@item
If you are running many @code{mysqld} servers on the same data on a
system that doesn't support good filesystem locks (normally handled by
the @code{lockd} daemon ) or if you are running
multiple servers with @code{--skip-locking}
@item
You have a crashed index/datafile that contains very wrong data that
got @code{mysqld} confused.
@item
You have found a bug in the data storage code. This isn't that likely,
but it's at least possible.  In this case you can try to change the file
type to another database handler by using @code{ALTER TABLE} on a
repaired copy of the table!
@end itemize

Because it is very difficult to know why something is crashing, first try to
check whether things that work for others crash for you.  Please try
the following things:

@itemize @bullet
@item
Cleaned up @code{NULL} handling for default values in @code{DESCRIBE
table_name}.
@item
Fixed @code{truncate()} to round up negative values to the nearest integer.
@item
Take down the @code{mysqld} daemon with @code{mysqladmin shutdown}, run
@code{myisamchk --silent --force */*.MYI} on all tables, and restart the
@code{mysqld} daemon.  This will ensure that you are running from a clean
state.  @xref{MySQL Database Administration}.

@item
Use @code{mysqld --log} and try to determine from the information in the log
whether some specific query kills the server. About 95% of all bugs are
related to a particular query!  Normally this is one of the last queries in
the log file just before MySQL restarted. @xref{Query log}.
If you can repeatadly kill MySQL with one of the queries, even
when you have checked all tables just before doing the query, then you
have been able to locate the bug and should do a bug report for this!
@xref{Bug reports}.

@item
Try to make a test case that we can use to reproduce the problem.
@xref{Reproduceable test case}.

@item
Try running the included mysql-test test and the MySQL
benchmarks.  @xref{MySQL test suite}.  They should test MySQL
rather well.  You can also add code that to the benchmarks to simulates
your application!  The benchmarks can be found in the @file{bench}
directory in the source distribution or, for a binary distribution, in
the @file{sql-bench} directory under your MySQL installation
directory.

@item
Try @code{fork_test.pl} and @code{fork2_test.pl}.

@item
If you configure MySQL for debugging, it will be much easier to
gather information about possible errors if something goes wrong.
Reconfigure MySQL with the @code{--with-debug} option or
@code{--with-debug=full} to @code{configure} and then recompile.
@xref{Debugging server}.

@item
Configuring MySQL for debugging causes a safe memory allocator to be
included that can find some errors. It also provides a lot of output about
what is happening.

@item
Have you applied the latest patches for your operating system?

@item
Use the @code{--skip-locking} option to @code{mysqld}.  On some systems, the
@code{lockd} lock manager does not work properly; the @code{--skip-locking}
option tells @code{mysqld} not to use external locking.  (This means that you
cannot run 2 @code{mysqld} servers on the same data and that you must be
careful if you use @code{myisamchk}, but it may be instructive to try the
option as a test.)

@item
Have you tried @code{mysqladmin -u root processlist} when @code{mysqld}
appears to be running but not responding?  Sometimes @code{mysqld} is not
comatose even though you might think so.  The problem may be that all
connections are in use, or there may be some internal lock problem.
@code{mysqladmin processlist} will usually be able to make a connection even
in these cases, and can provide useful information about the current number
of connections and their status.

@item
Run the command @code{mysqladmin -i 5 status} or @code{mysqladmin -i 5
-r status} or in a separate window to produce statistics while you run
your other queries.

@item
Try the following:
@enumerate
@item
Start @code{mysqld} from @code{gdb} (or in another debugger).
@xref{Using gdb on mysqld}.

@item
Run your test scripts.

@item
Print the backtrace and the local variables at the 3 lowest levels. In gdb you
can do this with the following commands when @code{mysqld} has crashed inside
gdb:

@example
backtrace
info local
up
info local
up
info local
@end example

With gdb you can also examine which threads exist with @code{info
threads} and switch to a specific thread with @code{thread #}, where
@code{#} is the thread id.
@end enumerate

@item
Try to simulate your application with a Perl script to force
MySQL to crash or misbehave.

@item
Send a normal bug report. @xref{Bug reports}. Be even more detailed
than usual.  Because MySQL works for many people, it may be that the
crash results from something that exists only on your computer (for example,
an error that is related to your particular system libraries).
@item
If you have a problem with tables with dynamic-length rows and you are
not using @code{BLOB/TEXT} columns (but only @code{VARCHAR} columns), you
can try to change all @code{VARCHAR} to @code{CHAR} with @code{ALTER
TABLE}.  This will force MySQL to use fixed-size rows.
Fixed-size rows take a little extra space, but are much more tolerant to
corruption!

The current dynamic row code has been in use at MySQL AB for at
least 3 years without any problems, but by nature dynamic-length rows are
more prone to errors, so it may be a good idea to try the above to see if
it helps!
@end itemize


@node Resetting permissions, Full disk, Crashing, Administration Issues
@appendixsubsec How to Reset a Forgotten Password

@cindex passwords, forgotten
@cindex passwords, resetting
@cindex root user, password resetting

If you have forgotten the @code{root} user password for MySQL, you
can restore it with the following procedure:

@enumerate
@item
Take down the @code{mysqld} server by sending a @code{kill} (not @code{kill
-9}) to the @code{mysqld} server.  The pid is stored in a @file{.pid}
file, which is normally in the MySQL database directory:

@example
kill `cat /mysql-data-directory/hostname.pid`
@end example

You must be either the Unix @code{root} user or the same user the server
runs as to do this.

@item
Restart @code{mysqld} with the @code{--skip-grant-tables} option.
@item
Connect to the @code{mysqld} server with @code{mysql -h hostname mysql} and change
the password with a @code{GRANT} command. @xref{GRANT,,@code{GRANT}}.
You can also do this with
@code{mysqladmin -h hostname -u user password 'new password'}
@item
Load the privilege tables with: @code{mysqladmin -h hostname
flush-privileges} or with the SQL command @code{FLUSH PRIVILEGES}.
@end enumerate

Note that after you started @code{mysqld} with @code{--skip-grant-tables},
any usage of @code{GRANT} commands will give you an @code{Unknown command}
error until you have executed @code{FLUSH PRIVILEGES}.


@node Full disk, Temporary files, Resetting permissions, Administration Issues
@appendixsubsec How MySQL Handles a Full Disk

@cindex full disk
@cindex disk full

@noindent
When a disk-full condition occurs, MySQL does the following:

@itemize @bullet
@item
It checks once every minute to see whether there is enough space to
write the current row. If there is enough space, it continues as if nothing had
happened.
@item
Every 6 minutes it writes an entry to the log file warning about the disk
full condition.
@end itemize

@noindent
To alleviate the problem, you can take the following actions:

@itemize @bullet
@item
To continue, you only have to free enough disk space to insert all records.
@item
To abort the thread, you must send a @code{mysqladmin kill} to the thread.
The thread will be aborted the next time it checks the disk (in 1 minute).
@item
Note that other threads may be waiting for the table that caused the disk
full condition.  If you have several ``locked'' threads, killing the one
thread that is waiting on the disk-full condition will allow the other
threads to continue.
@end itemize

Exceptions to the above behaveour is when you use @code{REPAIR} or
@code{OPTIMIZE} or when the indexes are created in a batch after an
@code{LOAD DATA INFILE} or after an @code{ALTER TABLE} statement.

All of the above commands may use big temporary files that left to
themself would cause big problems for the rest of the system.  If
MySQL gets disk full while doing any of the above operations,
it will remove the big temporary files and mark the table as crashed
(except for @code{ALTER TABLE}, in which the old table will be left
unchanged).


@node Temporary files, Problems with mysql.sock, Full disk, Administration Issues
@appendixsubsec Where MySQL Stores Temporary Files

MySQL uses the value of the @code{TMPDIR} environment variable as
the pathname of the directory in which to store temporary files.  If you don't
have @code{TMPDIR} set, MySQL uses the system default, which is
normally @file{/tmp} or @file{/usr/tmp}.  If the filesystem containing your
temporary file directory is too small, you should edit @code{safe_mysqld} to
set @code{TMPDIR} to point to a directory in a filesystem where you have
enough space!  You can also set the temporary directory using the
@code{--tmpdir} option to @code{mysqld}.

MySQL creates all temporary files as hidden files. This ensures
that the temporary files will be removed if @code{mysqld} is terminated.  The
disadvantage of using hidden files is that you will not see a big temporary
file that fills up the filesystem in which the temporary file directory is
located.

When sorting (@code{ORDER BY} or @code{GROUP BY}), MySQL normally
uses one or two temporary files. The maximum disk-space needed is:

@example
(length of what is sorted + sizeof(database pointer))
* number of matched rows
* 2
@end example

@code{sizeof(database pointer)} is usually 4, but may grow in the future for
really big tables.

For some @code{SELECT} queries, MySQL also creates temporary SQL
tables. These are not hidden and have names of the form @file{SQL_*}.

@code{ALTER TABLE} creates a temporary table in the same directory as
the original table.


@node Problems with mysql.sock, Timezone problems, Temporary files, Administration Issues
@appendixsubsec How to Protect or Change the MySQL Socket File @file{/tmp/mysql.sock}

@cindex @code{mysql.sock}, protection
@cindex deletion, @code{mysql.sock}

If you have problems with the fact that anyone can delete the
MySQL communication socket @file{/tmp/mysql.sock}, you can,
on most versions of Unix, protect your @file{/tmp} filesystem by setting
the @code{sticky} bit on it.  Log in as @code{root} and do the following:

@example
shell> chmod +t /tmp
@end example

This will protect your @file{/tmp} filesystem so that files can be deleted
only by their owners or the superuser (@code{root}).

You can check if the @code{sticky} bit is set by executing @code{ls -ld /tmp}.
If the last permission bit is @code{t}, the bit is set.

@cindex changing socket location

You can change the place where MySQL uses / puts the socket file the
following ways:

@itemize @bullet
@item
Specify the path in a global or local option file.
For example, put in @code{/etc/my.cnf}:

@example
[client]
socket=path-for-socket-file

[mysqld]
socket=path-for-socket-file
@end example
@xref{Option files}.
@item
Specifying this on the command-line to @code{safe_mysqld} and most
clients with the @code{--socket=path-for-socket-file} option.
@item
Specify the path to the socket in the @code{MYSQL_UNIX_PORT} environment
variable.
@item
Defining the path with the @code{configure} option
@code{--with-unix-socket-path=path-for-socket-file}. @xref{configure options}.
@end itemize

You can test that the socket works with this command:

@example
shell> mysqladmin --socket=/path/to/socket version
@end example

@node Timezone problems,  , Problems with mysql.sock, Administration Issues
@appendixsubsec Time Zone Problems

@cindex timezone problems
@cindex problems, timezone

@tindex TZ environment variable
@tindex Environment variable, TZ

If you have a problem with @code{SELECT NOW()} returning values in GMT and
not your local time, you have to set the @code{TZ} environment variable to
your current time zone.  This should be done for the environment in which
the server runs, for example, in @code{safe_mysqld} or @code{mysql.server}.
@xref{Environment variables}.


@node Query Issues, Table Definition Issues, Administration Issues, Problems
@appendixsec Query Related Issues


@menu
* Case sensitivity::            Case Sensitivity in Searches
* Using DATE::                  Problems Using @code{DATE} Columns
* Problems with NULL::          Problems with @code{NULL} Values
* Problems with alias::         Problems with @code{alias}
* Deleting from related tables::  Deleting Rows from Related Tables
* No matching rows::            Solving Problems with No Matching Rows
* Problems with float::         Problems with floating-point comparison
@end menu

@node Case sensitivity, Using DATE, Query Issues, Query Issues
@appendixsubsec Case Sensitivity in Searches

@cindex case sensitivity, in searches
@cindex searching, and case-sensitivity
@cindex Chinese
@cindex Big5 Chinese character encoding

By default, MySQL searches are case-insensitive (although there are
some character sets that are never case-insensitive, such as @code{czech}).
That means that if you search with @code{col_name LIKE 'a%'}, you will get all
column values that start with @code{A} or @code{a}. If you want to make this
search case-sensitive, use something like @code{INSTR(col_name, "A")=1} to
check a prefix. Or use @code{STRCMP(col_name, "A") = 0} if the column value
must be exactly @code{"A"}.

Simple comparison operations (@code{>=, >, = , < , <=}, sorting and
grouping) are based on each character's ``sort value''. Characters with
the same sort value (like E, e and é) are treated as the same character!

In older MySQL versions @code{LIKE} comparisons where done on
the uppercase value of each character (E == e but E <> é).  In newer
MySQL versions @code{LIKE} works just like the other comparison
operators.

If you want a column always to be treated in case-sensitive fashion,
declare it as @code{BINARY}. @xref{CREATE TABLE, , @code{CREATE TABLE}}.

If you are using Chinese data in the so-called big5 encoding, you want to
make all character columns @code{BINARY}. This works because the sorting
order of big5 encoding characters is based on the order of ASCII codes.


@node Using DATE, Problems with NULL, Case sensitivity, Query Issues
@appendixsubsec Problems Using @code{DATE} Columns

@findex DATE

@cindex DATE columns, problems
@cindex problems, @code{DATE} columns

The format of a @code{DATE} value is @code{'YYYY-MM-DD'}. According to ANSI
SQL, no other format is allowed. You should use this format in @code{UPDATE}
expressions and in the WHERE clause of @code{SELECT} statements.  For
example:

@example
mysql> SELECT * FROM tbl_name WHERE date >= '1997-05-05';
@end example

As a convenience, MySQL automatically converts a date to a number if
the date is used in a numeric context (and vice versa). It is also smart
enough to allow a ``relaxed'' string form when updating and in a @code{WHERE}
clause that compares a date to a @code{TIMESTAMP}, @code{DATE}, or a
@code{DATETIME} column.  (Relaxed form means that any punctuation character
may be used as the separator between parts. For example, @code{'1998-08-15'}
and @code{'1998#08#15'} are equivalent.) MySQL can also convert a
string containing no separators (such as @code{'19980815'}), provided it
makes sense as a date.

The special date @code{'0000-00-00'} can be stored and retrieved as
@code{'0000-00-00'.} When using a @code{'0000-00-00'} date through
@code{MyODBC}, it will automatically be converted to @code{NULL} in
@code{MyODBC} Version 2.50.12 and above, because ODBC can't handle this kind of
date.

Because MySQL performs the conversions described above, the following
statements work:

@example
mysql> INSERT INTO tbl_name (idate) VALUES (19970505);
mysql> INSERT INTO tbl_name (idate) VALUES ('19970505');
mysql> INSERT INTO tbl_name (idate) VALUES ('97-05-05');
mysql> INSERT INTO tbl_name (idate) VALUES ('1997.05.05');
mysql> INSERT INTO tbl_name (idate) VALUES ('1997 05 05');
mysql> INSERT INTO tbl_name (idate) VALUES ('0000-00-00');

mysql> SELECT idate FROM tbl_name WHERE idate >= '1997-05-05';
mysql> SELECT idate FROM tbl_name WHERE idate >= 19970505;
mysql> SELECT MOD(idate,100) FROM tbl_name WHERE idate >= 19970505;
mysql> SELECT idate FROM tbl_name WHERE idate >= '19970505';
@end example

@noindent
However, the following will not work:

@example
mysql> SELECT idate FROM tbl_name WHERE STRCMP(idate,'19970505')=0;
@end example

@code{STRCMP()} is a string function, so it converts @code{idate} to
a string and performs a string comparison.  It does not convert
@code{'19970505'} to a date and perform a date comparison.

Note that MySQL does no checking whether the date is
correct.  If you store an incorrect date, such as @code{'1998-2-31'}, the
wrong date will be stored. If the date cannot be converted to any reasonable
value, a @code{0} is stored in the @code{DATE} field.  This is mainly a speed
issue and we think it is up to the application to check the dates, and not
the server.


@node Problems with NULL, Problems with alias, Using DATE, Query Issues
@appendixsubsec Problems with @code{NULL} Values

@cindex @code{NULL} values, vs. empty values

@tindex NULL

The concept of the @code{NULL} value is a common source of confusion for
newcomers to SQL, who often think that @code{NULL} is the same thing as an
empty string @code{""}.  This is not the case! For example, the following
statements are completely different:

@example
mysql> INSERT INTO my_table (phone) VALUES (NULL);
mysql> INSERT INTO my_table (phone) VALUES ("");
@end example

Both statements insert a value into the @code{phone} column, but the first
inserts a @code{NULL} value and the second inserts an empty string.  The
meaning of the first can be regarded as ``phone number is not known'' and the
meaning of the second can be regarded as ``she has no phone''.

In SQL, the @code{NULL} value is always false in comparison to any
other value, even @code{NULL}.  An expression that contains @code{NULL}
always produces a @code{NULL} value unless otherwise indicated in
the documentation for the operators and functions involved in the
expression. All columns in the following example return @code{NULL}:

@example
mysql> SELECT NULL,1+NULL,CONCAT('Invisible',NULL);
@end example

If you want to search for column values that are @code{NULL}, you
cannot use the @code{=NULL} test. The following statement returns no
rows, because @code{expr = NULL} is FALSE, for any expression:

@example
mysql> SELECT * FROM my_table WHERE phone = NULL;
@end example

To look for @code{NULL} values, you must use the @code{IS NULL} test.
The following shows how to find the @code{NULL} phone number and the
empty phone number:

@example
mysql> SELECT * FROM my_table WHERE phone IS NULL;
mysql> SELECT * FROM my_table WHERE phone = "";
@end example

Note that you can only add an index on a column that can have @code{NULL}
values if you are using MySQL Version 3.23.2 or newer and are using the
@code{MyISAM} or @code{InnoDB} table type.
In earlier versions and with other table types, you must declare such
columns @code{NOT NULL}. This also means you cannot then insert
@code{NULL} into an indexed column.

@findex LOAD DATA INFILE
When reading data with @code{LOAD DATA INFILE}, empty columns are updated
with @code{''}. If you want a @code{NULL} value in a column, you should use
@code{\N} in the text file. The literal word @code{'NULL'} may also be used
under some circumstances.
@xref{LOAD DATA, , @code{LOAD DATA}}.

When using @code{ORDER BY}, @code{NULL} values are presented first. If you
sort in descending order using @code{DESC}, @code{NULL} values are presented
last.  When using @code{GROUP BY}, all @code{NULL} values are regarded as
equal.

To help with @code{NULL} handling, you can use the @code{IS NULL} and
@code{IS NOT NULL} operators and the @code{IFNULL()} function.

@cindex @code{TIMESTAMP}, and @code{NULL} values
@cindex @code{AUTO_INCREMENT}, and @code{NULL} values
@cindex @code{NULL} values, and @code{TIMESTAMP} columns
@cindex @code{NULL} values, and @code{AUTO_INCREMENT} columns
For some column types, @code{NULL} values are handled specially.  If you
insert @code{NULL} into the first @code{TIMESTAMP} column of a table, the
current date and time is inserted.  If you insert @code{NULL} into an
@code{AUTO_INCREMENT} column, the next number in the sequence is inserted.


@node Problems with alias, Deleting from related tables, Problems with NULL, Query Issues
@appendixsubsec Problems with @code{alias}

@tindex alias

You can use an alias to refer to a column in the @code{GROUP BY},
@code{ORDER BY}, or in the @code{HAVING} part.  Aliases can also be used
to give columns better names:

@example
SELECT SQRT(a*b) as rt FROM table_name GROUP BY rt HAVING rt > 0;
SELECT id,COUNT(*) AS cnt FROM table_name GROUP BY id HAVING cnt > 0;
SELECT id AS "Customer identity" FROM table_name;
@end example

Note that ANSI SQL doesn't allow you to refer to an alias in a
@code{WHERE} clause.  This is because when the @code{WHERE} code is
executed the column value may not yet be determined. For example, the
following query is @strong{illegal}:

@example
SELECT id,COUNT(*) AS cnt FROM table_name WHERE cnt > 0 GROUP BY id;
@end example

The @code{WHERE} statement is executed to determine which rows should
be included in the @code{GROUP BY} part while @code{HAVING} is used to
decide which rows from the result set should be used.


@node Deleting from related tables, No matching rows, Problems with alias, Query Issues
@appendixsubsec Deleting Rows from Related Tables

@cindex deleting, rows
@cindex rows, deleting
@cindex tables, deleting rows

As MySQL doesn't yet support sub-selects, nor the use of more than one
table in the @code{DELETE} statement (prior to Version 4.0), you should
use the following approach to delete rows from 2 related tables:

@enumerate
@item
@code{SELECT} the rows based on some @code{WHERE} condition in the main table.
@item
@code{DELETE} the rows in the main table based on the same condition.
@item
@code{DELETE FROM related_table WHERE related_column IN (selected_rows)}.
@end enumerate

If the total number of characters in the query with
@code{related_column} is more than 1,048,576 (the default value of
@code{max_allowed_packet}, you should split it into smaller parts and
execute multiple @code{DELETE} statements.  You will probably get the
fastest @code{DELETE} by only deleting 100-1000 @code{related_column}
id's per query if the @code{related_column} is an index.  If the
@code{related_column} isn't an index, the speed is independent of the
number of arguments in the @code{IN} clause.


@node No matching rows, Problems with float, Deleting from related tables, Query Issues
@appendixsubsec Solving Problems with No Matching Rows

@cindex no matching rows
@cindex rows, matching problems

If you have a complicated query that has many tables and that doesn't
return any rows, you should use the following procedure to find out what
is wrong with your query:

@enumerate
@item
Test the query with @code{EXPLAIN} and check if you can find something that is
obviously wrong. @xref{EXPLAIN, , @code{EXPLAIN}}.

@item
Select only those fields that are used in the @code{WHERE} clause.

@item
Remove one table at a time from the query until it returns some rows.
If the tables are big, it's a good idea to use @code{LIMIT 10} with the query.

@item
Do a @code{SELECT} for the column that should have matched a row against
the table that was last removed from the query.

@item
If you are comparing @code{FLOAT} or @code{DOUBLE} columns with numbers that
have decimals, you can't use @code{'='}. This problem is common in most
computer languages because floating-point values are not exact values.
In most cases, changing the @code{FLOAT} to a @code{DOUBLE} will fix this.
@xref{Problems with float}.

@item
If you still can't figure out what's wrong, create a minimal test that can
be run with @code{mysql test < query.sql} that shows your problems.
You can create a test file with @code{mysqldump --quick database tables > query.sql}. Open the file in an editor, remove some insert lines (if there are
too many of these), and add your select statement at the end of the file.

Test that you still have your problem by doing:

@example
shell> mysqladmin create test2
shell> mysql test2 < query.sql
@end example

Post the test file using @code{mysqlbug} to @email{mysql@@lists.mysql.com}.
@end enumerate

@node Problems with float,  , No matching rows, Query Issues
@appendixsubsec Problems with Floating-Point Comparison

floating-point numbers cause confusion sometimes, because these numbers
are not stored as exact values inside computer architecture. What one
can see on the screen usually is not the exact value of the number.

Field types @code{FLOAT}, @code{DOUBLE} and @code{DECIMAL} are such.

@example
CREATE TABLE t1 (i INT, d1 DECIMAL(9,2), d2 DECIMAL(9,2));
INSERT INTO t1 VALUES (1, 101.40, 21.40), (1, -80.00, 0.00),
(2, 0.00, 0.00), (2, -13.20, 0.00), (2, 59.60, 46.40),
(2, 30.40, 30.40), (3, 37.00, 7.40), (3, -29.60, 0.00),
(4, 60.00, 15.40), (4, -10.60, 0.00), (4, -34.00, 0.00),
(5, 33.00, 0.00), (5, -25.80, 0.00), (5, 0.00, 7.20),
(6, 0.00, 0.00), (6, -51.40, 0.00);

mysql> SELECT i, SUM(d1) AS a, SUM(d2) AS b
    -> FROM t1 GROUP BY i HAVING a <> b;
+------+--------+-------+
| i    | a      | b     |
+------+--------+-------+
|    1 |  21.40 | 21.40 |
|    2 |  76.80 | 76.80 |
|    3 |   7.40 |  7.40 |
|    4 |  15.40 | 15.40 |
|    5 |   7.20 |  7.20 |
|    6 | -51.40 |  0.00 |
+------+--------+-------+
@end example

The result is correct. Although the first five records look like they
shouldn't pass the comparison test, they may do so because the
difference between the numbers show up around tenth decimal, or so
depending on computer architecture.

The problem cannot be solved by using ROUND() (or similar function),
because the result is still a floating-point number. Example:

@example
mysql> SELECT i, ROUND(SUM(d1), 2) AS a, ROUND(SUM(d2), 2) AS b
    -> FROM t1 GROUP BY i HAVING a <> b;
+------+--------+-------+
| i    | a      | b     |
+------+--------+-------+
|    1 |  21.40 | 21.40 |
|    2 |  76.80 | 76.80 |
|    3 |   7.40 |  7.40 |
|    4 |  15.40 | 15.40 |
|    5 |   7.20 |  7.20 |
|    6 | -51.40 |  0.00 |
+------+--------+-------+
@end example

This is what the numbers in row 'a' look like:
@example
mysql> SELECT i, ROUND(SUM(d1), 2)*1.0000000000000000 AS a,
    -> ROUND(SUM(d2), 2) AS b FROM t1 GROUP BY i HAVING a <> b;
+------+----------------------+-------+
| i    | a                    | b     |
+------+----------------------+-------+
|    1 |  21.3999999999999986 | 21.40 |
|    2 |  76.7999999999999972 | 76.80 |
|    3 |   7.4000000000000004 |  7.40 |
|    4 |  15.4000000000000004 | 15.40 |
|    5 |   7.2000000000000002 |  7.20 |
|    6 | -51.3999999999999986 |  0.00 |
+------+----------------------+-------+
@end example

Depending on the computer architecture you may or may not see similar results.
Each CPU may evaluate floating-point numbers differently. For example in
some machines you may get 'right' results by multiplaying both arguments
with 1, an example follows.

@strong{WARNING: NEVER TRUST THIS METHOD IN YOUR APPLICATION, THIS IS
AN EXAMPLE OF A WRONG METHOD!!!}

@example
mysql> SELECT i, ROUND(SUM(d1), 2)*1 AS a, ROUND(SUM(d2), 2)*1 AS b
    -> FROM t1 GROUP BY i HAVING a <> b;
+------+--------+------+
| i    | a      | b    |
+------+--------+------+
|    6 | -51.40 | 0.00 |
+------+--------+------+
@end example

The reason why the above example seems to be working is that on the
particular machine where the test was done, the CPU floating-point
arithmetics happens to round the numbers to same, but there is no
rule that any CPU should do so, so it cannot be trusted.

The correct way to do floating-point number comparison is to first
decide on what is the wanted tolerance between the numbers and then do
the comparsion against the tolerance number. For example, if we agree on
that floating-point numbers should be regarded the same, if they are
same with precision of one of ten thousand (0.0001), the comparsion
should be done like this:

@example
mysql> SELECT i, SUM(d1) AS a, SUM(d2) AS b FROM t1
    -> GROUP BY i HAVING ABS(a - b) > 0.0001;
+------+--------+------+
| i    | a      | b    |
+------+--------+------+
|    6 | -51.40 | 0.00 |
+------+--------+------+
1 row in set (0.00 sec)
@end example

And vice versa, if we wanted to get rows where the numbers are the same,
the test would be:

@example
mysql> SELECT i, SUM(d1) AS a, SUM(d2) AS b FROM t1
    -> GROUP BY i HAVING ABS(a - b) < 0.0001;
+------+-------+-------+
| i    | a     | b     |
+------+-------+-------+
|    1 | 21.40 | 21.40 |
|    2 | 76.80 | 76.80 |
|    3 |  7.40 |  7.40 |
|    4 | 15.40 | 15.40 |
|    5 |  7.20 |  7.20 |
+------+-------+-------+
@end example

@node Table Definition Issues,  , Query Issues, Problems
@appendixsec Table Definition Related Issues


@menu
* ALTER TABLE problems::        Problems with @code{ALTER TABLE}.
* Change column order::         How To Change the Order of Columns in a Table
* Temporary table problems::    TEMPORARY TABLE problems
@end menu

@node ALTER TABLE problems, Change column order, Table Definition Issues, Table Definition Issues
@appendixsubsec Problems with @code{ALTER TABLE}.

@tindex ALTER TABLE

@code{ALTER TABLE} changes a table to the current character set.
If you during @code{ALTER TABLE} get a duplicate key error, then the cause
is either that the new character sets maps to keys to the same value
or that the table is corrupted, in which case you should run
@code{REPAIR TABLE} on the table.

If @code{ALTER TABLE} dies with an error like this:

@example
Error on rename of './database/name.frm' to './database/B-a.frm' (Errcode: 17)
@end example

the problem may be that MySQL has crashed in a previous @code{ALTER
TABLE} and there is an old table named @file{A-something} or
@file{B-something} lying around.  In this case, go to the MySQL data
directory and delete all files that have names starting with @code{A-} or
@code{B-}.  (You may want to move them elsewhere instead of deleting them.)

@code{ALTER TABLE} works the following way:

@itemize @bullet
@item Create a new table named @file{A-xxx} with the requested changes.
@item All rows from the old table are copied to @file{A-xxx}.
@item The old table is renamed @file{B-xxx}.
@item @file{A-xxx} is renamed to your old table name.
@item @file{B-xxx} is deleted.
@end itemize

If something goes wrong with the renaming operation, MySQL tries to
undo the changes.  If something goes seriously wrong (this shouldn't happen,
of course), MySQL may leave the old table as @file{B-xxx}, but a
simple rename on the system level should get your data back.


@node Change column order, Temporary table problems, ALTER TABLE problems, Table Definition Issues
@appendixsubsec How To Change the Order of Columns in a Table

@cindex reordering, columns
@cindex columns, changing
@cindex changing, column order
@cindex tables, changing column order

The whole point of SQL is to abstract the application from the data
storage format.  You should always specify the order in which you wish to
retrieve your data.  For example:

@example
SELECT col_name1, col_name2, col_name3 FROM tbl_name;
@end example

will return columns in the order @code{col_name1}, @code{col_name2}, @code{col_name3}, whereas:

@example
SELECT col_name1, col_name3, col_name2 FROM tbl_name;
@end example

will return columns in the order @code{col_name1}, @code{col_name3}, @code{col_name2}.

You should @strong{never}, in an application, use @code{SELECT *} and
retrieve the columns based on their position, because the order in which
columns are returned @strong{cannot} be guaranteed over time. A simple
change to your database may cause your application to fail rather
dramatically.

If you want to change the order of columns anyway, you can do it as follows:

@enumerate
@item
Create a new table with the columns in the right order.
@item
Execute
@code{INSERT INTO new_table SELECT fields-in-new_table-order FROM old_table}.
@item
Drop or rename @code{old_table}.
@item
@code{ALTER TABLE new_table RENAME old_table}.
@end enumerate


@node Temporary table problems,  , Change column order, Table Definition Issues
@appendixsubsec TEMPORARY TABLE problems

@cindex temporary tables, problems

The following are a list of the limitations with @code{TEMPORARY TABLES}.

@itemize @bullet
@item
A temporary table can only be of type @code{HEAP}, @code{ISAM},
@code{MyISAM} or @code{InnoDB}.
@item
You can't use temporary tables more than once in the same query.
For example, the following doesn't work.

@example
mysql> SELECT * FROM temporary_table, temporary_table AS t2;
@end example

We plan to fix the above in 4.0.
@item
You can't use @code{RENAME} on a @code{TEMPORARY} table.
Note that @code{ALTER TABLE org_name RENAME new_name} works!

We plan to fix the above in 4.0.
@end itemize




@node Contrib, Credits, Problems, Top
@appendix Contributed Programs

@cindex contributed programs
@cindex programs, contributed

Many users of MySQL have contributed @emph{very} useful support
tools and add-ons.

A list of what is available at
@uref{http://www.mysql.com/Downloads/Contrib/}
(or any mirror) is shown here.

Please visit our Software Portal at
@uref{http://www.mysql.com/portal/software/}.
The community facilities there also allow for your input!

If you want to build MySQL support for the Perl @code{DBI}/@code{DBD}
interface, you should fetch the @code{Data-Dumper}, @code{DBI}, and
@code{Msql-Mysql-modules} files and install them.
@xref{Perl support}.

Note: The programs listed here can be freely downloaded and used.
They are copyrighted by their respective owners.
Please see individual product documentation for more details on licensing and terms.
MySQL AB assumes no liability for the correctness of the information in this
chapter or for the proper operation of the programs listed herein.

@appendixsec APIs

@cindex Perl, modules
@itemize @bullet
@item Perl Modules
@itemize @minus
@item @uref{http://www.mysql.com/Downloads/Contrib/Data-Dumper-2.101.tar.gz}
Perl @code{Data-Dumper} module.  Useful with @code{DBI}/@code{DBD} support for
older Perl installations.

@item @uref{http://www.mysql.com/Downloads/Contrib/DBI-1.18.tar.gz}
Perl @code{DBI} module.

@item @uref{http://www.mysql.com/Downloads/Contrib/KAMXbase1.2.tar.gz}
Convert between @file{.dbf} files and MySQL tables.  Perl
module written by Pratap Pereira @email{pereira@@ee.eng.ohio-state.edu},
extended by Kevin A. McGrail @email{kmcgrail@@digital1.peregrinehw.com}.
This converter can handle MEMO fields.

@item @uref{http://www.mysql.com/Downloads/Contrib/Msql-Mysql-modules-1.2218.tar.gz}
Perl @code{DBD} module to access mSQL and MySQL databases.

@item @uref{http://www.mysql.com/Downloads/Contrib/Data-ShowTable-3.3.tar.gz}
Perl @code{Data-ShowTable} module.  Useful with @code{DBI}/@code{DBD} support.

@item @uref{http://www.mysql.com/Downloads/Contrib/HandySQL-1.1.tar.gz}
HandySQL is a MySQL access module. It offers a C interface embedded in Perl and is
approximately 20% faster than regular DBI.

@end itemize

@cindex JDBC
@item JDBC
@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/Contrib/mm.mysql.jdbc-1.2c.tar.gz}
The mm JDBC driver for MySQL. This is a production release
and is actively developed. By Mark Matthews
(@email{mmatthew@@ecn.purdue.edu}).

@item @uref{http://www.mysql.com/Downloads/Contrib/mm.mysql.jdbc-2.0pre5.tar.gz}
The mm JDBC driver for MySQL. This is a pre-release beta version
and is actively developed. By Mark Matthews
(@email{mmatthew@@ecn.purdue.edu}).
The two drivers above have an LGPL
license. Please check @uref{http://www.worldserver.com/mm.mysql/} for
the latest drivers (and other JDBC information) because these drivers may be out of date.

@item @uref{http://www.caucho.com/projects/jdbc-mysql/index.xtp}
The Resin commercial JDBC driver, which is released under @code{Open Source}.
It claims to be faster than the mm driver, but we haven't received that much
information about this yet.

@item @uref{http://www.mysql.com/Downloads/Contrib/twz1jdbcForMysql-1.0.4-GA.tar.gz}
The twz driver: A type 4 JDBC driver by Terrence W. Zellers
@email{zellert@@voicenet.com}.  This is commercial but is free for
private and educational use. (Not supported anymore.)
@c no answer from server 990830
@c You can always find the latest driver at @uref{http://www.voicenet.com/~zellert/tjFM/}.
@item @uref{http://www.mysql.com/Downloads/Contrib/pmdamysql.tgz}
A MySQL PMDA. Provides MySQL server status and configuration
variables.
@end itemize

@cindex OLEDB
@item OLEDB
@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/Win32/MyOLEDB3.exe}
MyOLEDB 3.0 installation package from SWSoft.
@item @uref{http://www.mysql.com/Downloads/Win32/mysql-oledb-3.0.0.zip}
Source for MyOLEDB 3.0.
@item @uref{http://www.mysql.com/Downloads/Win32/MySamples.zip}
Examples and documentation for MyOLEDB.
@item @uref{http://www.mysql.com/Downloads/Win32/MyOLEDB.chm}
Help files for MyOLEDB.
@item @uref{http://www.mysql.com/Downloads/Win32/libmyodbc.zip}
Static MyODBC library used for build MyOLEDB. Based on MyODBC code.
@end itemize

@cindex C++
@item C++
@itemize @bullet

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql-c++-0.02.tar.gz}
MySQL C++ wrapper library. By Roland Haenel,
@email{rh@@ginster.net}.

@item @uref{http://www.mysql.com/Downloads/Contrib/MyDAO.tar.gz}
MySQL C++ API. By Satish @email{spitfire@@pn3.vsnl.net.in}. Inspired
by Roland Haenel's C++ API and Ed Carp's MyC library.

@item @uref{http://www.mysql.com/download_mysql++.html}
MySQL C++ API (more than just a wrapper library). Originally by
@email{kevina@@clark.net}. Nowadays maintained by Sinisa at MySQL AB.

@item @uref{http://nelsonjr.homepage.com/NJrAPI/}
A C++ database independent library that supports MySQL.
@end itemize

@cindex Delphi
@item Delphi
@itemize @bullet

@item @uref{http://www.mysql.com/Downloads/Contrib/DelphiMySQL2.zip}
Delphi interface to @code{libmysql.dll}, by @email{bsilva@@umesd.k12.or.us}.

@item @uref{http://www.mysql.com/Downloads/Contrib/Udmysql.pas}
A wrapper for libmysql.dll for usage in Delphi. By Reiner Sombrowsky.

@item @uref{http://www.fichtner.net/delphi/mysql.delphi.phtml}
A Delphi Interface to MySQL, with source code. By Matthias Fichtner.

@item @uref{http://www.productivity.org/projects/tmysql/}
@code{TmySQL}, a library to use MySQL with Delphi.

@item @uref{https://sourceforge.net/projects/zeoslib/}
Zeos Library is a set of delphi native datasets and database
components for MySql, PostgreSql, Interbase, MS SQL, Oracle,
DB/2. Also it includes development tools such as Database
Explorer and Database Designer.

@item @uref{http://www.mysql.com/Downloads/Contrib/Win32/SBMySQL50Share.exe}
Delphi 5 Shareware MySQL Dataset Components.
@end itemize

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql-ruby-2.2.0.tar.gz}
MySQL Ruby module. By TOMITA Masahiro @email{tommy@@tmtm.org}
Ruby is an Object-Oriented Interpreter Language
(@uref{http://www.netlab.co.jp/ruby/}).

@item @uref{http://www.mysql.com/Downloads/Contrib/JdmMysqlDriver-0.1.0.tar.gz}
A VisualWorks 3.0 Smalltalk driver for MySQL. By
@email{joshmiller@@earthlink.net}.

@item @uref{http://www.mysql.com/Downloads/Contrib/Db.py}
Python module with caching. By @email{gandalf@@rosmail.com}.

@item @uref{http://www.mysql.com/Downloads/Contrib/MySQLmodule-1.4.tar.gz}
Python interface for MySQL. By Joseph Skinner @email{joe@@earthlight.co.nz}. Modified by Joerg Senekowitsch @email{senekow@@ibm.net}.

@item @uref{http://www.mysql.com/Downloads/Contrib/MySQL-python-0.3.0.tar.gz}
MySQLdb Python is an DB-API v2.0-compliant interface to MySQL. Transactions
are supported if the server and tables support them.  It is
thread-safe, and contains a compatibility module for older code
written for the no-longer-maintained MySQLmodule interface.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql_mex_12.tar.gz}
An interface program for the Matlab program by MathWorks.  The interface
is done by Kimmo Uutela and John Fisher (not by Mathworks).
Check @uref{http://boojum.hut.fi/~kuutela/mysqlmex.html}
for more information.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysqltcl-1.53.tar.gz}
Tcl interface for MySQL. Based on @file{msqltcl-1.50.tar.gz}.
Updated by Tobias Ritzau, @email{tobri@@ida.liu.se}.

@item @uref{http://www.mysql.com/Downloads/Contrib/MyC-0.1.tar.gz}
A Visual Basic-like API, by Ed Carp.

@item @uref{http://www.mysql.com/Downloads/Contrib/Vdb-dflts-2.1.tar.gz}
This is a new version of a set of library utilities intended
to provide a generic interface to SQL database engines such that your
application becomes a 3-tiered application.  The advantage is that you
can easily switch between and move to other database engines by
implementing one file for the new backend without making any
changes to your applications. By @email{damian@@cablenet.net}.

@item @uref{http://www.mysql.com/Downloads/Contrib/DbFramework-1.10.tar.gz}
DbFramework is a collection of classes for manipulating MySQL
databases.  The classes are loosely based on the CDIF Data Model
Subject Area. By Paul Sharpe @email{paul@@miraclefish.com}.

@item @uref{http://www.mysql.com/Downloads/Contrib/pike-mysql-1.4.tar.gz}
MySQL module for pike. For use with the Roxen web server.

@item @uref{http://www.mysql.com/Downloads/Contrib/squile.tar.gz}
Module for @code{guile} that allows @code{guile} to interact with SQL
databases. By Hal Roberts.

@item @uref{http://www.mysql.com/Downloads/Contrib/stk-mysql.tar.gz}
Interface for Stk. Stk is the Tk widgets with Scheme underneath instead of Tcl.
By Terry Jones.

@item @uref{http://www.mysql.com/Downloads/Contrib/eiffel-wrapper-1.0.tar.gz}
Eiffel wrapper by Michael Ravits.

@item @uref{http://www.mysql.com/Downloads/Contrib/SQLmy0.06.tgz}
FlagShip Replaceable Database Driver (RDD) for MySQL. By Alejandro
Fernandez Herrero.
The Flagship RDD homepage is at @uref{http://www.fship.com/rdds.html}.

@item @uref{http://www.mysql.com/Downloads/Contrib/mydsn-1.0.zip}
Binary and source for @code{mydsn.dll}. mydsn should be used to build
and remove the DSN registry file for the MyODBC driver in Coldfusion
applications.  By Miguel Angel Solórzano.

@item @uref{http://www.mysql.com/Downloads/Contrib/MySQL-ADA95_API.zip}
An ADA95 interface to the MySQL API. By Francois Fabien.

@item @uref{http://www.mysql.com/Downloads/Contrib/MyTool-DLL_for_VB_and_MySQL.zip}
A DLL with MySQL C API for Visual Basic.
By Ken Menzel @email{kenm@@icarz.com}.

@item @uref{http://www.mysql.com/Downloads/Contrib/MYSQLX.EXE}
MySQL ActiveX Object for directly accessing your MySQL
servers from IIS/ASP, VB, VC++ skipping the slower ODBC methods. Fully
updatable, multi-threaded with full support for all MySQL fieldtypes
(version 2001.1.1). By SciBit @uref{http://www.scibit.com/}.

@item @uref{http://www.fastflow.it/mylua/}
MyLUA home page; how to use the LUA language to write MySQL
@code{PROCEDURE} that can be loaded runtime.

@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/Contrib/lua-4.0.tar.gz}
LUA 4.0
@item @uref{http://www.mysql.com/Downloads/Contrib/mylua-3.23.32.1.tar.gz}
Patch for MySQL 3.23.32 to use LUA 4.0.  By Cristian Giussani.
@end itemize

@item @uref{http://www.mysql.com/Downloads/Contrib/patched_myodbc.zip}
Patch (for Omniform 4.0 support) to the MyODBC driver.
By Thomas Thaele @email{tthaele@@papenmeier.de}

@end itemize

@appendixsec  Clients

@itemize @bullet
@item Graphical clients
@itemize @minus
@item @uref{http://www.ideit.com/products/dbvis/}
DbVisualizer, a freeware JDBC client to graphically visualise
the data and structure of several databases simultaneously.
By Innovative-IT Development AB.

@item @uref{http://www.mysql.com/downloads/gui-clients.html}
MySQLGUI, the MySQL GUI client homepage. By Sinisa at MySQL AB.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql_navigator_0.9.0.tar.gz}
MySQL Navigator is a MySQL database server GUI client program, distributed
under GPL license. The purpose of MySQL Navigator is to provide a useful
client interface to MySQL database servers, whilst supporting multiple
operating systems and languages. You can currently import/export database,
enter queries, get result sets, edit scripts, run scripts, add, alter, and
delete users, and retrieve client and server information. Uses QT 2.2.
The homepage for MySQL Navigator is at @uref{http://sql.kldp.org/mysql/}.

@item @uref{http://www.mysql.com/Downloads/Win32/secman.zip}
A user and security management GUI for MySQL on Windows.
By Martin Jeremic.
The homepage for MySQL Security GUI is at @uref{http://jsoft.webjump.com/}.

@item @uref{http://www.mysql.com/Downloads/Contrib/kmysqladmin-0.4.1.tar.gz}.
@item @uref{http://www.mysql.com/Downloads/Contrib/kmysqladmin-0.4.1-1.src.rpm}.
@item @uref{http://www.mysql.com/Downloads/Contrib/kmysqladmin-0.4.1-1.i386.rpm}
An administration tool for the MySQL server using QT / KDE.  Tested
only on Linux.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql-admin-using-java+swing.tar.gz}
Java client using Swing, by Fredy Fischer (@email{se-afs@@dial.eunet.ch}).
You can always find the latest version at
@uref{http://www.trash.net/~ffischer/admin/}.

@item @uref{http://www.mysql.com/Downloads/Win32/MySQL-Maker-1.0.zip}.
Shareware MySQL client for Windows. It's a WYSIWYG tool which allows
you to create, change and delete databases and tables.
You can change field - structure and add, change and delete data in
these tables directly without ODBC-driver.
The MySQL Maker homepage is at @uref{http://www.presult.de/presult/frames/fs_mysqlmaker.html}.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysqlwinadmn.zip}
Windows GUI (binary only) to administrate a database, by David B. Mansel,
@email{david@@zhadum.org}.

@item @uref{http://home.online.no/~runeberg/myqa/}
MyQA is a Linux-based query client for the MySQL database server. MyQA
lets you enter SQL queries, execute them, and view the results, all in a
graphical user interface. The GUI is roughly similar to that of the
'Query Analyzer' client that comes with MS SQL Server.

@item @uref{http://www.opex.atnet.ru/mysqlmanager/}
MySQL Manager is a graphical MySQL server manager for MySQL server
written in Java.

@item @uref{http://www.mysql.com/Downloads/Win32/netadmin.zip}
An administrator tool for MySQL on Windows 95/98 and Windows NT
4.0. Only tested with MySQL Versions 3.23.5 - 3.23.7. Written
using the Tmysql components.

You can write queries and show tables, indexes, table syntax, and
administrate user, host, and database and so on. This is beta and
still has some bugs. You can test the program with all features. Please
send bugs and hints to Marco Suess @email{ms@@it-netservice.de}. Original
URL @uref{http://www.it-netservice.de/pages/software/}.

@item @uref{http://www.mysql.com/Downloads/Win32/netadmin2.zip}
New version of netadmin. See above for details.

@item @uref{http://www.mysql.com/Downloads/Win32/ARTADMIN203.EXE}
Atronic's MySQL client for Windows 2.0.3.0.
The home page for this can be found at @uref{http://www.artronic.hr/}.

@item @uref{http://www.mysql.com/Downloads/Win32/W9xstop.zip}
Utility from Artronic to stop MySQL on win9x.

@item @uref{http://bardo.hyperlink.cz/mysqlmon/}
A light weight GUI client for Windows.

@item @uref{http://www.mysqlfront.de/}
MySQLfront is a very nice Windows client with lots of useful features.
By Angsar Becker.

@item @uref{http://www.dbtools.com.br/}
Dbtools, a tool to manage MySQL databases. Currently only for Windows.
Some features:
@itemize @bullet
@item Manage servers, databases, tables, columns, indexes, and users
@item Import wizard to import structure and data from MS Access, MS Excel, Dbase, FoxPro, Paradox, and ODBC Databases.

@item @uref{http://www.mysql.com/Downloads/Contrib/KMYENG113.zip}
An administrator GUI for MySQL. Works only on Windows, no source.
Available in English and Japanese. By Mitunobu Kaneko.
Home page: @uref{http://sql.jnts.ne.jp/}
@end itemize

@item @uref{http://www.mysql.com/Downloads/Contrib/xmysqladmin-1.0.tar.gz}
An X-based front end to the MySQL database engine.  It allows reloads,
status check, process control, myisamchk, grant/revoke privileges,
creating databases, dropping databases, create, alter, browse, and drop
tables.  Originally by Gilbert Therrien, @email{gilbert@@ican.net} but
now in public domain and supported by MySQL AB.

@item @uref{http://www.mysql.com/Downloads/Contrib/xmysql-1.9.tar.gz}.
xmysqlA front end to the MySQL database engine. It allows for simple
queries and table maintenance, as well as batch queries. By Rick
Mehalick, @email{dblhack@@wt.net}.
The xmysql homepage is at @uref{http://web.wt.net/~dblhack/}
Requires @uref{http://bragg.phys.uwm.edu/xforms/} (xforms 0.88) to work.

@item @uref{http://www.tamos.net/sw/dbMetrix/}
dbMatrix is an @code{Open Source} client for exploring databases and executing
SQL. Supports MySQL, Oracle, PostgreSQL, and mSQL.

@item @uref{http://www.multimania.com/bbrox/GtkSQL/}
GtkSQL is a query tool for MySQL and PostgreSQL.

@item @uref{http://dbman.linux.cz/}
dbMan is a query tool written in Perl. Uses DBI and Tk.

@item @uref{http://www.mysql.com/Downloads/Win32/Msc201.EXE} (Mascon 202)
@item @uref{http://www.mysql.com/Downloads/Win32/FrMsc202.EXE} (Free Mascon 202)
Mascon is a powerful Win32 GUI for the administering MySQL server
databases. Mascon's features include visual table design, connections to
multiple servers, data and blob editing of tables, security setting, SQL
color coding, dump functionality and much more.
The Mascon homepage is at @uref{http://www.scibit.com/Products/Software/Utils/Mascon.asp}.
@item @uref{http://www.virtualbeer.net/dbui/}
DBUI is a Gtk graphical database editor.
@item @uref{http://www.rtlabs.com/}
MacSQL is a GUI for MySQL, ODBC, and JDBC databases for the Mac OS.
@item @uref{http://www.caleb.com.au/}
JRetriever is a generic database front-end tool for JDBC compliant
databases written with Java 2. JRetriever displays database
tables/views in a Windows explorer-like front end. Users can retrieve
data either by clicking on the table folder or by composing their own SQL
statements with our built-in SQL editor. The tool has been tested with
Oracle 8 and MySQL as the back-end databases. It requires JDK 1.3 from
JavaSoft.
@item @uref{http://www.jetools.com/products/databrowser/}
The DataBrowser is a cross-database, cross-platform data access tool. It is more
user friendly than tools like SQL Plus, psql (command-line based tools). It is more
flexible than TOAD, ISQL, PGAccess which are GUI's that are limitied to a single
platform or database.
@item @uref{http://www.intrex.net/amit/software/}
The SQL Console is a standalone java application that allows you to connect to a
 SQL database system and issue SQL queries and updates. It has an easy-to use
graphical user interface. The SQL Console uses JDBC to connect to the database
systems and, therefore, with proper JDBC drivers, you can use this utility to
connect to some of the most popular database systems.
@item @uref{http://www.mysql.com/Downloads/Contrib/mysql_mmc.zip}
MySQL MMC is a GUI Management Tool developed using kdevelop
with a very good interface completely like Microsoft
Enterprise Tool (for SQL Server) or Sybase Central. We
can use it to manage server, database, table, index,
users and to edit table data in grid or execute Sql
by Query Analysis.
@end itemize

@cindex Web clients
@item Web Clients
@itemize @minus
@item @uref{http://www.mysql.com/Downloads/Contrib/mysqladmin-atif-1.0.tar.gz}
WWW MySQL administrator for the @code{user,} @code{db} and
@code{host} tables.  By Tim Sailer, modified by Atif Ghaffar
@email{aghaffar@@artemedia.ch}.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql-webadmin-1.0a8-rz.tar.gz}
A tool written in PHP-FI to administrate MySQL databases
remotely over the web within a web-browser. By Peter Kuppelwieser,
@email{peter.kuppelwieser@@kantea.it}. Updated by Wim Bonis,
@email{bonis@@kiss.de}. Not maintained anymore!

@item @uref{http://www.mysql.com/Downloads/Contrib/mysqladm.tar.gz}
MySQL Web Database Administration written in Perl.  By Tim Sailer.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysqladm-2.tar.gz}
Updated version of @file{mysqladm.tar.gz}, by High Tide.

@item @uref{http://www.mysql.com/Downloads/Contrib/billowmysql.zip}
Updated version of @file{mysqladm.tar.gz}, by Ying Gao.  You can get the
newest version from @uref{http://civeng.com/sqldemo/} (the home site).

@item @uref{http://www.mysql.com/Downloads/Contrib/myadmin-0.4.tar.gz}.
MyAdmin is a web-based MySQL administrator by Mike Machado.
TheMyAdmin homepage is at @uref{http://myadmin.cheapnet.net/}

@item @uref{http://www.mysql.com/Downloads/Contrib/phpMyAdmin_2.2.0.tar.gz}
A set of PHP3-scripts to adminstrate MySQL over the WWW.

@item @uref{http://www.phpwizard.net/projects/phpMyAdmin/}
phpMyAdmin is a PHP3 tool in the spirit of mysql-webadmin, by Tobias Ratschiller, tobias@@dnet.it.

@item @uref{http://www.mysql.com/Downloads/Contrib/useradm.tar.gz}
MySQL administrator in PHP. By Ofni Thomas
@email{othomas@@vaidsystems.com}.

@item @uref{http://gossamer-threads.com/perl/mysqlman/mysql.cgi}
MySQLMan has similar functionality to phpmyadmin, but written with
Perl and using html templates. By Alex Krohn.
@end itemize

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql-editor.tar.gz}
This cgi scripts in Perl enables you to edit content of Mysql
database. By Tomas Zeman.

@item @uref{http://worldcommunity.com/opensource/futuresql/}
FutureSQL by Peter F. Brown, is a free, @code{Open Source} rapid application
development web database administration tool, written in Perl,
using MySQL. It uses @code{DBI:DBD} and @code{CGI.pm}.

FutureSQL allows one to easily set up config files to view, edit, delete,
and otherwise process records from a MySQL database. It uses a data
dictionary, configuration files and templates, and allows "pre-processing"
and "post-processing" on both fields, records, and operations.
@end itemize

@cindex web tools
@cindex tools,, web
@appendixsec Web Tools

@itemize @bullet

@item @uref{http://www.mysql.com/Downloads/Contrib/mod_mysql_include_1.0.tar.gz}
Apache module to include HTML from MySQL queries into your pages,
and run update queries.  Originally written to implement a simple fast
low-overhead banner-rotation system.  By Sasha Pachev.

@item @uref{http://htcheck.sourceforge.net/}
htCheck is a URL checker with
MySQL backend. Spidered URLs can later be queried using SQL to retrieve
various kinds of information, eg. broken links. Written by Gabriele Bartolini.

@item @uref{http://www.odbsoft.com/cook/sources.htm}
This package has various functions for generating html code from a SQL
table structure and for generating SQL statements (Select, Insert,
Update, Delete) from an html form. You can build a complete forms
interface to a SQL database (query, add, update, delete) without any
programming!  By Marc Beneteau, @email{marc@@odbsoft.com}.

@item @uref{http://www.mysql.com/Downloads/Contrib/sqlhtml.tar.gz}
SQL/HTML is an HTML database manager for MySQL using @code{DBI} 1.06.

@item @uref{http://www.mysql.com/Downloads/Contrib/udmsearch-3.0.23.tar.gz} (UdmSearch 3.0.23, stable version).
@item @uref{http://www.mysql.com/Downloads/Contrib/mnogosearch-3.1.12.tar.gz} (mnogosearch 3.1.12, development but recommended version)
@item @uref{http://search.mnoGo.ru/} (UdmSearch home page)
A SQL-based search engine for Internet. By
Alexander I. Barkov @email{bar@@izhcom.ru}.

@item @uref{http://www.mysql.com/Downloads/Contrib/wmtcl.doc}
@item @uref{http://www.mysql.com/Downloads/Contrib/wmtcl.lex}
With this you can write HTML files with inclusions of Tcl code. By
@email{vvs@@scil.npi.msu.su}.

@item @uref{http://www.mysql.com/Downloads/Contrib/www-sql-0.5.7.lsm}
@item @uref{http://www.mysql.com/Downloads/Contrib/www-sql-0.5.7.tar.gz}
A CGI program that parses an HTML file containing special tags, parses
them, and inserts data from a MySQL database.

@item @uref{http://www.mysql.com/Downloads/Contrib/genquery.zip}
Perl SQL database interface package for html.

@item @uref{http://www.mysql.com/Downloads/Contrib/cgi++-0.8.tar.gz}
A macro-processor to simply writing CGI/Database programs in C++ by Sasha Pachev.

@item  @uref{http://www.mysql.com/Downloads/Contrib/webboard-1.0.zip}
WebBoard 1.0, EU-Industries Internet-Message-Board.

@item @uref{http://www.mysql.com/Downloads/Contrib/DBIx-TextIndex-0.02.tar.gz}
Full-text searching with Perl on @code{BLOB}/@code{TEXT} columns by Daniel Koch.
@end itemize

@cindex tools, benchmarking
@cindex benchmarking, tools
@appendixsec Performance Benchmarking Tools

@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/super-smack/super-smack-1.0.tar.gz}
super-smack is a multi-threaded benchmarking tool for MySQL and
@strong{PostgreSQL}. Written in C++. Easy to extend to support other
databases that have C/C++ client libraries. By Sasha Pachev.
@end itemize

@cindex tools, authentication
@cindex authentication tools
@appendixsec Authentication Tools

@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/Contrib/ascend-radius-mysql-0.7.2.patch.gz}
This is an authentication and logging patch using MySQL for
Ascend-Radius. By @email{takeshi@@SoftAgency.co.jp}.

@item @uref{http://www.mysql.com/Downloads/Contrib/icradius-0.10.tar.gz} (icradius 0.10)
@uref{http://www.mysql.com/Downloads/Contrib/icradius.README} (icradius readme)

@item @uref{http://www.mysql.com/Downloads/Contrib/checkpassword-0.81-mysql-0.6.6.patch.gz}
MySQL authentication patch for QMAIL and checkpassword. These are
useful for management user (mail, pop account) by MySQL.
By @email{takeshi@@SoftAgency.co.jp}.

@item @uref{http://www.mysql.com/Downloads/Contrib/jradius-diff.gz}
MySQL support for Livingston's Radius 2.01. Authentication and
Accounting. By Jose de Leon, @email{jdl@@thevision.net}.

@item @uref{http://www.mysql.com/Downloads/Contrib/mod_auth_mysql-2.20.tar.gz}
Apache authentication module for MySQL. By Zeev Suraski,
@email{bourbon@@netvision.net.il}.

@c @strong{Please} register this module at:
@c @uref{http://bourbon.netvision.net.il/mysql/mod_auth_mysql/register.html}. The
@c registering information is only used for statistical purposes and will
@c encourage further development of this module!

@item @uref{http://www.mysql.com/Downloads/Contrib/mypasswd-2.0.tar.gz}
Extra for @code{mod_auth_mysql}. This is a little tool that allows you
to add/change user records storing group and/or password entries in
MySQL tables. By Harry Brueckner, @email{brueckner@@respublica.de}.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql-passwd.README}
@item @uref{http://www.mysql.com/Downloads/Contrib/mysql-passwd-1.2.tar.gz}
Extra for @code{mod_auth_mysql}. This is a two-part system for use with
@code{mod_auth_mysql}.

@item @uref{http://www.mysql.com/Downloads/Contrib/pam_mysql.tar.gz}
This module authenticates users via @code{pam}, using MySQL.

@item @uref{http://www.mysql.com/Downloads/Contrib/nsapi_auth_mysql.tar}
Netscape Web Server API (NSAPI) functions to authenticate (BASIC) users
against MySQL tables. By Yuan John Jiang.

@item @uref{http://www.mysql.com/Downloads/Contrib/qmail-1.03-mysql-0.6.6.patch.gz}
Patch for qmail to authenticate users from a MySQL table.
By @email{takeshi@@SoftAgency.co.jp}.

@item @uref{http://www.mysql.com/Downloads/Contrib/proftpd-1.2.0rc2-fix-mysql.patch}
Patch for proftpd1.2.0rc2. By @email{takeshi@@SoftAgency.co.jp}.

@item @uref{http://www.mysql.com/Downloads/Contrib/pwcheck_mysql-0.1.tar.gz}
An authentication module for the Cyrus IMAP server. By Aaron Newsome.


@end itemize

@cindex converters
@appendixsec Converters

@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/Contrib/mssql2mysql.txt}
Converter from MS-SQL to MySQL. By Michael Kofler.
The mssql2mysql home page is at
@uref{http://www.kofler.cc/mysql/mssql2mysql.html}.

@item @uref{http://www.mysql.com/Downloads/Contrib/dbf2mysql-1.14.tar.gz}
Convert between @file{.dbf} files and MySQL tables. By Maarten
Boekhold (@email{boekhold@@cindy.et.tudelft.nl}), William Volkman, and
Michael Widenius.  This converter includes rudimentary read-only support
for MEMO fields.

@item @uref{http://www.mysql.com/Downloads/Contrib/dbf2mysql-1.13.tgz}
Convert between @file{.dbf} files and MySQL tables. By Maarten
Boekhold, @email{boekhold@@cindy.et.tudelft.nl}, and Michael Widenius.
This converter can't handle MEMO fields.

@item @uref{http://www.mysql.com/Downloads/Contrib/dbf2mysql.zip}
Convert between FoxPro @file{.dbf} files and MySQL tables on Windows.
By Alexander Eltsyn, @email{ae@@nica.ru} or @email{ae@@usa.net}.

@item @uref{http://www.mysql.com/Downloads/Contrib/dbf2sql.zip}
Short and simple prg that can help you transport your data from foxpro
table into MySQL table. By Danko Josic.

@item @uref{http://www.mysql.com/Downloads/Contrib/dump2h-1.20.gz}
Convert from @code{mysqldump} output to a C header file. By Harry Brueckner,
@email{brueckner@@mail.respublica.de}.

@item @uref{http://www.mysql.com/Downloads/Contrib/exportsql.txt}
A script that is similar to @code{access_to_mysql.txt}, except that this
one is fully configurable, has better type conversion (including
detection of @code{TIMESTAMP} fields), provides warnings and suggestions
while converting, quotes @strong{all} special characters in text and
binary data, and so on.  It will also convert to @code{mSQL} v1 and v2,
and is free of charge for anyone. See
@uref{http://www.cynergi.net/exportsql/} for the latest version. By
Pedro Freire, @email{support@@cynergi.net}.  Note: Doesn't work with
Access2!

@item @uref{http://www.mysql.com/Downloads/Contrib/access_to_mysql.txt}
Paste this function into an Access module of a database that has the
tables you want to export. See also @code{exportsql}. By Brian Andrews.
Note: Doesn't work with Access2!

@item @uref{http://www.mysql.com/Downloads/Contrib/importsql.txt}
A script that does the exact reverse of @code{exportsql.txt}.  That is,
it imports data from MySQL into an Access database via
ODBC. This is very handy when combined with exportsql, because it lets you
use Access for all DB design and administration, and synchronise with
your actual MySQL server either way. Free of charge. See
@uref{http://www.netdive.com/freebies/importsql/} for any updates.
Created by Laurent Bossavit of NetDIVE.
@strong{Note}: doesn't work with Access2!

@item @uref{http://www.mysql.com/Downloads/Contrib/mdb2sql.bas}
Converter from Access97 to MySQL by Moshe Gurvich.

@item @uref{http://www.mysql.com/Downloads/Contrib/msql2mysqlWrapper-1.0.tgz}
A C wrapper from @code{mSQL} to MySQL. By @email{alfred@@sb.net}

@item @uref{http://www.mysql.com/Downloads/Contrib/sqlconv.pl}
A simple script that can be used to copy fields from one MySQL table
to another in bulk. Basically, you can run @code{mysqldump} and pipe it to
the @code{sqlconv.pl} script. The script will parse through the
@code{mysqldump} output and will rearrange the fields so they can be
inserted into a new table. An example is when you want to create a new
table for a different site you are working on, but the table is just a
bit different (that is - fields in different order, etc.).
By Steve Shreeve.
@item @uref{http://www.mysql.com/Downloads/Contrib/oracledump}
Perl program to convert Oracle databases to MySQL. Has same
output format as mysqldump. By Johan Andersson.

@item @uref{http://www.mysql.com/Downloads/Contrib/excel2mysql.pl}
Perl program to import Excel spreadsheets into a MySQL database. By Stephen Hurd @email{shurd@@sk.sympatico.ca}

@item @uref{http://www.mysql.com/Downloads/Contrib/T2S_100.ZIP}.
Windows program to convert text files to MySQL databases. By
Asaf Azulay.

@end itemize

@appendixsec Using MySQL with Other Products

@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/Contrib/emacs-sql-mode.tar.gz}
Raw port of a SQL mode for XEmacs. Supports completion. Original by
Peter D. Pezaris @email{pez@@atlantic2.sbi.com} and partial
MySQL port by David Axmark.

@item @uref{http://www.mysql.com/Downloads/Win32/myaccess97_1_4.zip} (MyAccess97 1.4)
@item @uref{http://www.mysql.com/Downloads/Win32/myaccess2000_1_4.zip} (MyAccess2000 1.4)

MyAccess is an AddIn for MS Access 97/2000 that allows you to manage
MySQL databases from within Access. Main functions are:
@itemize @minus
@item Create/Modify Tables
@item Execute Queries against MySQL
@item Extract ''Create Table-Scripts'' from MySQL
@item Import/Export tables from Access to MySQL and vice versa
@item Log Changes
@item Show a "Database Definition Report
@end itemize

Written by Hubertus Hiden.
The MyAccess homepage is at @uref{http://www.accessmysql.com/}.

@item @uref{http://www.mysql.com/Downloads/Contrib/radius-0.3.tar.gz}
Patches for @code{radiusd} to make it support MySQL. By Wim Bonis,
@email{bonis@@kiss.de}.
@end itemize

@cindex utilities
@appendixsec Utilities

@itemize @bullet
@item @uref{http://worldcommunity.com/opensource/utilities/mysql_backup.html}
MySQL Backup is a backup script for MySQL. By Peter F. Brown.

@item @uref{http://www.mysql.com/Downloads/Contrib/mytop}
@item @uref{http://public.yahoo.com/~jzawodn/mytop/} (mytop home page)
mytop is a Perl program that allows you to monitor MySQL servers by
viewing active threads, queries, and overall server performance
numbers. By Jeremy D. Zawodny.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql_watchdog.pl}
Monitor the MySQL daemon for possible lockups. By Yermo Lamers,
@email{yml@@yml.com}.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysqltop.tar.gz}
Sends a query in a fixed time interval to the server and shows the
resulting table. By Thomas Wana.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql_structure_dumper.tar.gz}
@item @uref{http://www.mysql.com/Downloads/Contrib/mysql_structure_dumper.tgz}
Prints the structure of every table in a database. By Thomas Wana.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysqlsync}.
A Perl script to keep remote copies of a MySQL database in sync with a
central master copy. By Mark Jeftovic.  @email{markjr@@easydns.com}.

@item @uref{http://www.mysql.com/Downloads/Contrib/MySQLTutor-0.2.tar.gz}.
MySQLTutor. A MySQL tutorial for beginners.

@item @uref{http://www.mysql.com/Downloads/Contrib/MySQLDB.zip}
@item @uref{http://www.mysql.com/Downloads/Contrib/MySQLDB-readme.html}.
A COM library for MySQL by Alok Singh.

@item @uref{http://www.mysql.com/Downloads/Contrib/mysql_replicate.pl}
Perl program that handles replication.  By @email{elble@@icculus.nsg.nwu.edu}

@item @uref{http://www.mysql.com/Downloads/Contrib/DBIx-TextIndex-0.02.tar.gz}
Perl script that uses reverse indexing to handle text searching.
By Daniel Koch.

@item @uref{http://www.mysql.com/Downloads/Contrib/dbcheck}
Perl script that takes a backup of tables before running isamchk on them.
By Elizabeth.

@item @uref{http://www.mysql.com/Downloads/Contrib/mybackup}.
@item @uref{http://www.mswanson.com/mybackup} (mybackup home page)
Wrapper for mysqldump to backup all databases. By Marc Swanson.

@item @uref{http://www.mysql.com/Downloads/Contrib/mdu.pl.gz}
Prints the storage usage of a MySQL database.
@end itemize

@cindex RPMs, for common tools
@cindex tools, RPMs for
@appendixsec RPMs for Common Tools (Most Are for RedHat 6.1)

@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/Contrib/perl-Data-ShowTable-3.3-2.i386.rpm}
@item @uref{http://www.mysql.com/Downloads/Contrib/perl-Msql-Mysql-modules-1.2210-2.i386.rpm}
@item @uref{http://www.mysql.com/Downloads/Contrib/php-pg-3.0.13-1.i386.rpm}
@item @uref{http://www.mysql.com/Downloads/Contrib/php-pg-manual-3.0.13-1.i386.rpm}
@item @uref{http://www.mysql.com/Downloads/Contrib/php-pg-mysql-3.0.13-1.i386.rpm}
@item @uref{http://www.mysql.com/Downloads/Contrib/phpMyAdmin-2.0.5-1.noarch.rpm}
@end itemize

@cindex functions, useful
@appendixsec Useful Functions
@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/Contrib/mysnprintf.c}
sprintf() function for SQL queries that can escape blobs. By Chunhua Liu.
@end itemize

@appendixsec Windows Programs
@itemize @bullet
@item @uref{http://www.mysql.com/Downloads/Contrib/LaunchMySQL.zip}
The program launches the MySQL server, shuts it down, and
display status information. By Bill Thompson
@end itemize

@appendixsec Uncategorised

@itemize @bullet

@item @uref{http://www.mysql.com/Downloads/Contrib/findres.pl}
Find reserved words in tables. By Nem W Schlecht.

@item @uref{http://www.mysql.com/Downloads/Contrib/handicap.tar.gz}
Performance handicapping system for yachts. Uses PHP. By
@email{rhill@@stobyn.ml.org}.

@item @uref{http://www.mysql.com/Downloads/Contrib/hylalog-1.0.tar.gz}
Store @code{hylafax} outgoing faxes in a MySQL database. By Sinisa
Milivojevic, @email{sinisa@@mysql.com}.

@item @uref{http://www.mysql.com/Downloads/Contrib/mrtg-mysql-1.0.tar.gz}
MySQL status plotting with MRTG, by Luuk de Boer, @email{luuk@@wxs.nl}.

@item @uref{http://www.mysql.com/Downloads/Contrib/wuftpd-2.4.2.18-mysql_support.2.tar.gz}
Patches to add logging to MySQL for WU-ftpd. By Zeev Suraski,
@email{bourbon@@netvision.net.il}.

@item @uref{http://www.mysql.com/Downloads/Contrib/wu-ftpd-2.6.0-mysql.4.tar.gz}
Patches to add logging to MySQL for WU-ftpd 2.6.0. By
@email{takeshi@@SoftAgency.co.jp}, based on Zeev Suraski wuftpd patches.

@item @uref{http://www.mysql.com/Downloads/Contrib/Old-Versions}
Previous versions of things found here that you probably won't be
interested in.
@end itemize
@page




@node Credits, News, Contrib, Top
@appendix Credits

@cindex developers, list of

This appendix lists the developers, contributors, and supporters that have
helped to make MySQL what it is today.

@menu
* Developers::                  Developers at MySQL AB
* Contributors::                Contributors to MySQL
* Supporters::                  Supporters to MySQL
@end menu


@node Developers, Contributors, Credits, Credits
@appendixsec Developers at MySQL AB

These are the developers that are or have been employed by @code{MySQL AB}
to work on the @code{MySQL} database software, roughly in the order they
started to work with us. Following each developer is a small list of the
tasks that the developer is responsible for, or the accomplishments they
have made.

@table @asis
@item Michael (Monty) Widenius
Has written the following parts of the MySQL database software:
@itemize @bullet
@item
All the main code in @code{mysqld}.
@item
New functions for the string library.
@item
Most of the @code{mysys} library.
@item
The @code{ISAM} and @code{MyISAM} libraries (B-tree index file
handlers with index compression and different record formats).
@item
The @code{HEAP} library. A memory table system with our superior full dynamic
hashing. In use since 1981 and published around 1984.
@item
The @code{replace} program (take a look at it, it's @strong{COOL}!).
@item
@code{MyODBC}, the ODBC driver for Windows95.
@item
Fixing bugs in MIT-pthreads to get it to work for MySQL Server. And
also Unireg, a curses-based application tool with many utilities.
@item
Porting of @code{mSQL} tools like @code{msqlperl}, @code{DBD}/@code{DBI}, and
@code{DB2mysql}.
@item
Most of crash-me and the foundation for the MySQL benchmarks.
@end itemize

@item David Axmark
@itemize @bullet
@item
Coordinator and initial main writer of the @strong{Reference Manual},
including enhancements to @code{texi2html}.
@item
Automatic web site updating from the manual.
@item
Initial Autoconf, Automake, and Libtool support.
@item
Licensing.
@item
Parts of all the text files.  (Nowadays only the @file{README} is
left. The rest ended up in the manual.)
@item
Lots of testing of new features.
@item
Our in-house Free Software legal expert.
@item
Mailing list maintainer (who never has the time to do it right...).
@item
Our original portability code (more than 10 years old now). Nowadays
only some parts of @code{mysys} are left.
@item
Someone for Monty to call in the middle of the night when he just got
that new feature to work.
@end itemize

@item Jani Tolonen
@itemize @bullet
@item
@code{mysqlimport}
@item
A lot of extensions to the @code{mysql} client.
@item
@code{PROCEDURE ANALYSE()}
@end itemize

@item Sinisa Milivojevic
@itemize @bullet
@item
Compression (with @code{zlib}) in the client/server protocol.
@item
Perfect hashing for the lexical analyser phase.
@item
Multi-row @code{INSERT}
@item
@code{mysqldump} -e option
@item
@code{LOAD DATA INFILE LOCAL}
@item
@code{SQL_CALC_FOUND_ROWS} @code{SELECT} option
@item
@code{--max-user-connections=...} option
@item
@code{net_read} and @code{net_write_timeout}
@item
@code{GRANT}/@code{REVOKE} and @code{SHOW GRANTS FOR}
@item
New client-server protocol for 4.0
@item
@code{UNION}.
@item
Multi-table @code{DELETE}/@code{UPDATE}
@item
The @code{MySQLGUI} client.
@item
Maintainer of @code{MySQL++}.
@end itemize

@item Tonu Samuel
@itemize @bullet
@item
Our security expert.
@item
Vio interface (the foundation for the encrypted client/server protocol).
@item
MySQL Filesystem (a way to use MySQL databases as files
and directories).
@item
The CASE expression.
@item
The MD5() and COALESCE() functions.
@item
@code{RAID} support for @code{MyISAM} tables.
@end itemize

@item Sasha Pachev
@itemize @bullet
@item
Replication.
@item
@code{SHOW CREATE TABLE}.
@item
mod_mysql_include
@item
cgi++
@item
mysql-bench
@end itemize

@item Matt Wagner
@itemize @bullet
@item
MySQL test suite.
@item
Our webmaster.
@end itemize

@item Miguel Solorzano
@itemize @bullet
@item
Win32 development.
@item
Winmysqladmin.
@end itemize

@item Timothy Smith
@itemize @bullet
@item
Dynamic character support.
@item
Responsible for MySQL configure.
@end itemize

@item Sergei Golubchik
@itemize @bullet
@item
Full-text search.
@item
Added keys to the @code{MERGE} library.
@end itemize

@item Jeremy Cole
@itemize @bullet
@item
Proofreading and editing this fine manual.
@item
@code{ALTER TABLE ... ORDER BY ...}.
@item
@code{UPDATE ... ORDER BY ...}.
@item
@code{DELETE ... ORDER BY ...}.
@end itemize

@item Indrek Siitan
@itemize @bullet
@item
Designer/programmer of our web interface.
@end itemize

@item Jorge del Conde
@itemize @bullet
@item
@code{MyCC} @code{MySQL Control Center}.
@item
Web portals.
@item
Win32 development.
@end itemize
@end table


@node Contributors, Supporters, Developers, Credits
@appendixsec Contributors to MySQL

@cindex contributors, list of

While @code{MySQL AB} owns all copyrights in the @code{MySQL server}
and the @code{MySQL manual}, we wish to recognise those who have made
contributions of one kind or another to the @code{MySQL distribution}.
Contributors are listed here, in somewhat random order:

@table @asis
@item Paul DuBois
Help with making the Reference Manual correct and understandable. That
includes rewriting Monty's and David's attempts at English into English
as other people know it.
@item Gianmassimo Vigazzola @email{qwerg@@mbox.vol.it} or @email{qwerg@@tin.it}
The initial port to Win32/NT.
@item Kim Aldale
Helped to rewrite Monty's and David's early attempts at English into
English.
@item Per Eric Olsson
For more or less constructive criticism and real testing of the dynamic
record format.
@item Irena Pancirov @email{irena@@mail.yacc.it}
Win32 port with Borland compiler.  @code{mysqlshutdown.exe} and
@code{mysqlwatch.exe}
@item David J. Hughes
For the effort to make a shareware SQL database. At TcX, the predecessor
of MySQL AB, we started with @code{mSQL}, but found that it couldn't
satisfy our purposes so instead we wrote a SQL interface to our
application builder Unireg.  @code{mysqladmin} and @code{mysql} client are
programs that were largely influenced by their @code{mSQL} counterparts.
We have put a lot of effort into making the MySQL syntax a superset of
@code{mSQL}. Many of the API's ideas are borrowed from @code{mSQL} to
make it easy to port free @code{mSQL} programs to the MySQL API.
The MySQL software doesn't contain any code from @code{mSQL}.
Two files in the distribution (@file{client/insert_test.c} and
@file{client/select_test.c}) are based on the corresponding (non-copyrighted)
files in the @code{mSQL} distribution, but are modified as examples showing
the changes necessary to convert code from @code{mSQL} to MySQL Server.
(@code{mSQL} is copyrighted David J. Hughes.)
@item Fred Fish
For his excellent C debugging and trace library. Monty has made a number
of smaller improvements to the library (speed and additional options).
@item Richard A. O'Keefe
For his public domain string library.
@item Henry Spencer
For his regex library, used in @code{WHERE column REGEXP regexp}.
@item Free Software Foundation
From whom we got an excellent compiler (@code{gcc}), the @code{libc} library
(from which we have borrowed @file{strto.c} to get some code working in Linux),
and the @code{readline} library (for the @code{mysql} client).
@item Free Software Foundation & The XEmacs development team
For a really great editor/environment used by almost everybody at
MySQL AB/TcX/detron.
@item Patrick Lynch
For helping us acquire @uref{http://www.mysql.com/}.
@item Fred Lindberg
For setting up qmail to handle the MySQL mailing list and for the
incredible help we got in managing the MySQL mailing lists.
@item Igor Romanenko @email{igor@@frog.kiev.ua}
@code{mysqldump} (previously @code{msqldump}, but ported and enhanced by
Monty).
@item Yuri Dario
For keeping up and extending the MySQL OS/2 port.
@item Tim Bunce, Alligator Descartes
For the @code{DBD} (Perl) interface.
@item Tim Bunce
Author of @code{mysqlhotcopy}.
@item Andreas Koenig @email{a.koenig@@mind.de}
For the Perl interface for MySQL Server.
@item Eugene Chan @email{eugene@@acenet.com.sg}
For porting PHP for MySQL Server.
@item Michael J. Miller Jr. @email{mke@@terrapin.turbolift.com}
For the first MySQL manual. And a lot of spelling/language
fixes for the FAQ (that turned into the MySQL manual a long
time ago).
@item Yan Cailin
First translator of the MySQL Reference Manual into simplified
chinese in early 2000 on which the Big5 and HK coded
(@uref{http://mysql.hitstar.com/}) versions were
based. @uref{http://linuxdb.yeah.net, Personal home page at
linuxdb.yeah.net}.
@item Giovanni Maruzzelli @email{maruzz@@matrice.it}
For porting iODBC (Unix ODBC).
@item Chris Provenzano
Portable user level pthreads. From the copyright: This product includes
software developed by Chris Provenzano, the University of California,
Berkeley, and contributors.  We are currently using version 1_60_beta6
patched by Monty (see @file{mit-pthreads/Changes-mysql}).
@item Xavier Leroy @email{Xavier.Leroy@@inria.fr}
The author of LinuxThreads (used by the MySQL Server on Linux).
@item Zarko Mocnik @email{zarko.mocnik@@dem.si}
Sorting for Slovenian language and the @file{cset.tar.gz} module that makes
it easier to add other character sets.
@item "TAMITO" @email{tommy@@valley.ne.jp}
The @code{_MB} character set macros and the ujis and sjis character sets.
@item Joshua Chamas @email{joshua@@chamas.com}
Base for concurrent insert, extended date syntax, debugging on NT, and
answering on the MySQL mailing list.
@item Yves Carlier @email{Yves.Carlier@@rug.ac.be}
@code{mysqlaccess}, a program to show the access rights for a user.
@item Rhys Jones @email{rhys@@wales.com} (And GWE Technologies Limited)
For the JDBC, a module to extract data from a MySQL Database with a Java client.
@item Dr Xiaokun Kelvin ZHU @email{X.Zhu@@brad.ac.uk}
Further development of the JDBC driver and other MySQL-related
Java tools.
@item James Cooper @email{pixel@@organic.com}
For setting up a searchable mailing list archive at his site.
@item Rick Mehalick @email{Rick_Mehalick@@i-o.com}
For @code{xmysql}, a graphical X client for MySQL Server.
@item Doug Sisk @email{sisk@@wix.com}
For providing RPM packages of MySQL for RedHat Linux.
@item Diemand Alexander V. @email{axeld@@vial.ethz.ch}
For providing RPM packages of MySQL for RedHat Linux-Alpha.
@item Antoni Pamies Olive @email{toni@@readysoft.es}
For providing RPM versions of a lot of MySQL clients for Intel
and SPARC.
@item Jay Bloodworth @email{jay@@pathways.sde.state.sc.us}
For providing RPM versions for MySQL Version 3.21.
@item Jochen Wiedmann @email{wiedmann@@neckar-alb.de}
For maintaining the Perl @code{DBD::mysql} module.
@item Therrien Gilbert @email{gilbert@@ican.net}, Jean-Marc Pouyot @email{jmp@@scalaire.fr}
French error messages.
@item Petr snajdr, @email{snajdr@@pvt.net}
Czech error messages.
@item Jaroslaw Lewandowski @email{jotel@@itnet.com.pl}
Polish error messages.
@item Miguel Angel Fernandez Roiz
Spanish error messages.
@item Roy-Magne Mo @email{rmo@@www.hivolda.no}
Norwegian error messages and testing of Version 3.21.#.
@item Timur I. Bakeyev @email{root@@timur.tatarstan.ru}
Russian error messages.
@item @email{brenno@@dewinter.com} & Filippo Grassilli @email{phil@@hyppo.com}
Italian error messages.
@item Dirk Munzinger @email{dirk@@trinity.saar.de}
German error messages.
@item Billik Stefan @email{billik@@sun.uniag.sk}
Slovak error messages.
@item Stefan Saroiu @email{tzoompy@@cs.washington.edu}
Romanian error messages.
@item Peter Feher
Hungarian error messages.
@item Roberto M. Serqueira
Portugise error messages.
@item Carsten H. Pedersen
Danish error messages
@item David Sacerdote @email{davids@@secnet.com}
Ideas for secure checking of DNS hostnames.
@item Wei-Jou Chen @email{jou@@nematic.ieo.nctu.edu.tw}
Some support for Chinese(BIG5) characters.
@item  Wei He @email{hewei@@mail.ied.ac.cn}
A lot of functionality for the Chinese(GBK) character set.
@item Zeev Suraski @email{bourbon@@netvision.net.il}
@code{FROM_UNIXTIME()} time formatting, @code{ENCRYPT()} functions, and
@code{bison} advisor.
Active mailing list member.
@item Luuk de Boer @email{luuk@@wxs.nl}
Ported (and extended) the benchmark suite to @code{DBI}/@code{DBD}. Have
been of great help with @code{crash-me} and running benchmarks. Some new
date functions. The @code{mysql_setpermissions} script.
@item Jay Flaherty @email{fty@@mediapulse.com}
Big parts of the Perl @code{DBI}/@code{DBD} section in the manual.
@item Paul Southworth @email{pauls@@etext.org}, Ray Loyzaga @email{yar@@cs.su.oz.au}
Proof-reading of the Reference Manual.
@item Alexis Mikhailov @email{root@@medinf.chuvashia.su}
User-definable functions (UDFs); @code{CREATE FUNCTION} and
@code{DROP FUNCTION}.
@item Andreas F. Bobak @email{bobak@@relog.ch}
The @code{AGGREGATE} extension to UDF functions.
@item Ross Wakelin @email{R.Wakelin@@march.co.uk}
Help to set up InstallShield for MySQL-Win32.
@item Jethro Wright III @email{jetman@@li.net}
The @file{libmysql.dll} library.
@item James Pereria @email{jpereira@@iafrica.com}
Mysqlmanager, a Win32 GUI tool for administrating MySQL Server.
@item Curt Sampson @email{cjs@@portal.ca}
Porting of MIT-pthreads to NetBSD/Alpha and NetBSD 1.3/i386.
@item Antony T. Curtis @email{antony.curtis@@olcs.net}
Porting of the MySQL Database software to OS/2.
@item Martin Ramsch @email{m.ramsch@@computer.org}
Examples in the MySQL Tutorial.
@item Steve Harvey
For making @code{mysqlaccess} more secure.
@item Konark IA-64 Centre of Persistent Systems Private Limited
@uref{http://www.pspl.co.in/konark/}.  Help with the Win64 port of the
MySQL server.
@item Albert Chin-A-Young.
Configure updates for Tru64, large file support and better TCP wrappers
support.
@item John Birrell
Emulation of @code{pthread_mutex()} for OS/2.
@item Benjamin Pflugmann
Extended @code{MERGE} tables to handle @code{INSERTS}. Active member
on the MySQL mailing lists.
@end table

Other contributors, bugfinders, and testers: James H. Thompson, Maurizio
Menghini, Wojciech Tryc, Luca Berra, Zarko Mocnik, Wim Bonis, Elmar
Haneke, @email{jehamby@@lightside}, @email{psmith@@BayNetworks.com},
@email{duane@@connect.com.au}, Ted Deppner @email{ted@@psyber.com},
Mike Simons, Jaakko Hyvatti.

And lots of bug report/patches from the folks on the mailing list.

A big tribute goes to those that help us answer questions on the
@code{mysql@@lists.mysql.com} mailing list:

@table @asis
@item Daniel Koch @email{dkoch@@amcity.com}
Irix setup.
@item Luuk de Boer @email{luuk@@wxs.nl}
Benchmark questions.
@item Tim Sailer @email{tps@@users.buoy.com}
@code{DBD-mysql} questions.
@item Boyd Lynn Gerber @email{gerberb@@zenez.com}
SCO-related questions.
@item Richard Mehalick @email{RM186061@@shellus.com}
@code{xmysql}-related questions and basic installation questions.
@item Zeev Suraski @email{bourbon@@netvision.net.il}
Apache module configuration questions (log & auth), PHP-related
questions, SQL syntax-related questions and other general questions.
@item Francesc Guasch @email{frankie@@citel.upc.es}
General questions.
@item Jonathan J Smith @email{jsmith@@wtp.net}
Questions pertaining to OS-specifics with Linux, SQL syntax, and other
things that might need some work.
@item David Sklar @email{sklar@@student.net}
Using MySQL from PHP and Perl.
@item Alistair MacDonald @email{A.MacDonald@@uel.ac.uk}
Not yet specified, but is flexible and can handle Linux and maybe HP-UX.
Will try to get user to use @code{mysqlbug}.
@item John Lyon @email{jlyon@@imag.net}
Questions about installing MySQL on Linux systems, using either
@file{.rpm} files or compiling from source.
@item Lorvid Ltd. @email{lorvid@@WOLFENET.com}
Simple billing/license/support/copyright issues.
@item Patrick Sherrill @email{patrick@@coconet.com}
ODBC and VisualC++ interface questions.
@item Randy Harmon @email{rjharmon@@uptimecomputers.com}
@code{DBD}, Linux, some SQL syntax questions.
@end table


@node Supporters,  , Contributors, Credits
@appendixsec Supporters to MySQL

@cindex contributing companies, list of

While @code{MySQL AB} owns all copyrights in the @code{MySQL server}
and the @code{MySQL manual}, we wish to recognise the following companies,
which helped us finance the development of the @code{MySQL server},
such as by paying us for developing a new feature or giving us hardware
for development of the @code{MySQL server}.

@table @asis
@item VA Linux / Andover.net
Funded replication.
@item NuSphere
Editing of the MySQL manual.
@item Stork Design studio
The MySQL web site in use between 1998-2000.
@item Intel
Contributed to development on Windows and Linux platforms.
@item Compaq
Contributed to Development on Linux/Alpha.
@item SWSoft
Development on the embedded @code{mysqld} version.
@item FutureQuest
@code{--skip-show-database}
@end table




@node News, Porting, Credits, Top
@appendix MySQL Change History

@cindex ChangeLog
@cindex changes, log
@cindex log, changes

This appendix lists the changes from version to version in the MySQL
source code.

Note that we tend to update the manual at the same time we make changes to
MySQL.  If you find a version listed here that you can't
find on the MySQL download page
(@uref{http://www.mysql.com/downloads/}),
this means that the version has not yet been released!

@menu
* News-4.0.x::                  Changes in release 4.0.x (Development; Alpha)
* News-3.23.x::                 Changes in release 3.23.x (Stable)
* News-3.22.x::                 Changes in release 3.22.x (Older; Still supported)
* News-3.21.x::                 Changes in release 3.21.x
* News-3.20.x::                 Changes in release 3.20.x
* News-3.19.x::                 Changes in release 3.19.x
@end menu


@node News-4.0.x, News-3.23.x, News, News
@appendixsec Changes in release 4.0.x (Development; Alpha)

We are now working actively on MySQL 4.0 and will only provide critical
bug fixes for MySQL 3.23.  We will update this section as we add new
features, so that others can follow our development.

Our TODO section contains what we plan to have in 4.0. @xref{TODO MySQL 4.0}.

@cindex changes, version 4.0

@menu
* News-4.0.2::                  Changes in release 4.0.2
* News-4.0.1::                  Changes in release 4.0.1 (23 Dec 2001)
* News-4.0.0::                  Changes in release 4.0.0 (Oct 2001: Alpha)
@end menu

@node News-4.0.2, News-4.0.1, News-4.0.x, News-4.0.x
@appendixsubsec Changes in release 4.0.2

@itemize @bullet
@item
Fixed bug in @code{GROUP BY} with two or more fields, where at least one
field can contain @code{NULL} values.
@item
Use @code{Turbo Boyer-Moore} algorithm to speed up @code{LIKE "%keyword%"}
searches.
@item
Fixed bug in @code{DROP DATABASE} with symlink.
@item
Fixed crash in @code{REPAIR ... USE_FRM}.
@item
Fixed bug in @code{EXPLAIN} with @code{LIMIT offset != 0}
@item
Fixed bug in phrase operator @code{"..."} in boolean full-text search.
@item
Fixed bug that caused duplicated rows when using truncation operator
@code{*} in boolean full-text search.
@item
Fixed bug in truncation operator of boolean full-text search (wrong results
when there are only @code{+word*}'s in the query).
@item
Fixed bug in boolean full-text search that caused a crash when an identical
@code{MATCH} expression that did not use an index appeared twice.
@item
Query cache is now automatically disabled in @code{mysqldump}.
@item
Fixed problem on Windows 98 that made sending of results very slow.
@item
Boolean full-text search weighting scheme changed to something more reasonable.
@item
Fixed bug in boolean full-text search that caused MySQL to ignore queries of
@code{ft_min_word_len} characters.
@item
Boolean full-text search now supports ``phrase searches''.
@item
New configure option @code{--without-query-cache}. 
@item
Memory allocation strategy for ``root memory'' changed. Block size now grows 
with number of allocated blocks.
@item
@code{INET_NTOA()} now returns @code{NULL} if you give it an argument that
is too large (greater than the value corresponding to @code{255.255.255.255}).
@item
Fix @code{SQL_CALC_FOUND_ROWS} to work with @code{UNION}s. It will work only
if the first @code{SELECT} has this option and if there is global @code{LIMIT}
for the entire statement. For the moment, this requires using parentheses for
individual @code{SELECT} queries within the statement.
@item
Don't give an error for @code{CREATE TABLE ...(... VARCHAR(0))}.
@item
Fixed @code{SIGINT} and @code{SIGQUIT} problems in @file{mysql.cc} on Linux
with some @code{glibc} versions.
@item
Fixed bug in @file{convert.cc}, which is caused by having an incorrect
@code{net_store_length()} linked in the @code{CONVERT::store()} method.
@item
@code{DOUBLE} and @code{FLOAT} columns now honor the @code{UNSIGNED} flag
on storage.
@item
@code{InnoDB} now retains foreign key constraints through @code{ALTER TABLE}
and @code{CREATE/DROP INDEX}.
@item
@code{InnoDB} now allows foreign key constraints to be added through the
@code{ALTER TABLE} syntax.
@item
@code{InnoDB} tables can now be set to automatically grow in size (autoextend).
@item
Added @code{--ignore-lines=n} option to @code{mysqlimport}.  This has the
same effect as the @code{IGNORE n LINES} clause for @code{LOAD DATA}.
@item
Fixed bug in @code{UNION} with last offset being transposed to total result
set.
@item
@code{REPAIR ... USE_FRM} added.
@item
Fixed that @code{DEFAULT_SELECT_LIMIT} is always imposed on @code{UNION}
result set.
@item
Fixed that some @code{SELECT} options can appear only in the first
@code{SELECT}.
@item
Fixed bug with @code{LIMIT} with @code{UNION}, where last select is in
the braces.
@item
Fixed that full-text works fine with @code{UNION} operations.
@item
Fixed bug with indexless boolean full-text search.
@item
Fixed bug that sometimes appeared when full-text search was used
with ``const'' tables.
@item
Fixed incorrect error value when doing a @code{SELECT} with an empty
@code{HEAP} table.
@item
Use @code{ORDER BY column DESC} now sorts @code{NULL} values first.
@item
Fixed bug in @code{WHERE key_name='constant' ORDER BY key_name DESC}.
@item
Fixed bug in @code{SELECT DISTINCT ... ORDER BY DESC} optimisation.
@item
Fixed bug in @code{... HAVING 'GROUP_FUNCTION'(xxx) IS [NOT] NULL}.
@item
Fixed bug in truncation operator for boolean full-text search.
@item
Allow value of @code{--user=#} option for @code{mysqld} to be specified
as a numeric user ID.
@item
Fixed that @code{SELECT 0 LIMIT 0} doesn't hang thread.
@item
Fixed some problems with @code{USE/IGNORE INDEX} when using
many keys with the same start column.
@item
Don't use table scan with @code{BerkeleyDB} and @code{InnoDB} tables when
we can use an index that covers the whole row.
@item
Optimised @code{InnoDB} sort-buffer handling to take less memory.
@item
Fixed bug in multi-table @code{DELETE} and @code{InnoDB} tables.
@item
Fixed problem with @code{TRUNCATE} and @code{InnoDB} tables that produced the
error @code{Can't execute the given command because you have active locked
tables or an active transaction}.
@item
Added @code{NO_UNSIGNED_SUBTRACTION} to the set of flags that may be
specified with the @code{--sql-mode} option for @code{mysqld}.  It disables
unsigned arithmetic rules when it comes to subtraction. (This will make
MySQL 4.0 behave more closely to 3.23 with @code{UNSIGNED} columns).
@item
Added @code{WITH MAX_QUERIES_PER_HOUR=#} to @code{GRANT} command.
@item
The result returned for all bit functions (@code{|}, @code{<<}, ...) is now of
type @code{unsigned integer}.
@item
Added detection of @code{nan} values in @code{MyISAM} to make it possible to
repair tables with @code{nan} in float or double columns.
@item
Fixed new bug in @code{myisamchk} where it didn't correctly update number of
``parts'' in the @code{MyISAM} index file.
@item
Changed to use @code{autoconf} 2.52 (from @code{autoconf} 2.13).
@item
Fixed optimisation problem where the MySQL Server was in ``preparing'' state
for a long time when selecting from an empty table which had contained
a lot of rows.
@item
Fixed bug in complicated join with @code{const} tables. This fix also
improves performance a bit when referring to another table from a
@code{const} table.
@item
First pre-version of multi-table @code{UPDATE} statement.
@item
Fixed bug in multi-table @code{DELETE}.
@item
Fixed bug in @code{SELECT CONCAT(argument_list) ... GROUP BY 1}.
@item
@code{INSERT ... SELECT} did a full rollback in case of an error. Fixed
so that we only roll back the last statement in the current transaction.
@item
Fixed bug with empty expression for boolean full-text search.
@item
Fixed core dump bug in updating full-text key from/to @code{NULL}.
@item
ODBC compatibility: Added @code{BIT_LENGTH()} function.
@item
Fixed core dump bug in @code{GROUP BY BINARY column}.
@item
Added support for @code{NULL} keys in @code{HEAP} tables.
@item
Use index for @code{ORDER BY} in queries of type:
@code{SELECT * FROM t WHERE key_part1=1 ORDER BY key_part1 DESC,key_part2 DESC}
@item
Fixed bug in @code{FLUSH QUERY CACHE}.
@item
Added @code{CAST()} and @code{CONVERT()} functions. The @code{CAST} and
@code{CONVERT} functions are nearly identical and mainly useful when you
want to create a column with a specific type in a @code{CREATE ... SELECT}
statement.  For more information, read @ref{Cast Functions}.
@item
@code{CREATE ... SELECT} on @code{DATE} and @code{TIME} functions now
create columns of the expected type.
@item
Changed order in which keys are created in tables.
@item
Added new columns @code{Null} and @code{Index_type} to @code{SHOW INDEX}
output.
@item
Added @code{--no-beep} and @code{--prompt} options to @code{mysql} command-line client.
@item
New feature: management of user resources.
@example
GRANT ... WITH MAX_QUERIES_PER_HOUR = N1
               MAX_UPDATES_PER_HOUR = N2
               MAX_CONNECTIONS_PER_HOUR = N3;
@end example
@xref{User resources}.
@end itemize

@node News-4.0.1, News-4.0.0, News-4.0.2, News-4.0.x
@appendixsubsec Changes in release 4.0.1 (23 Dec 2001)

@itemize @bullet
@item
Fixed bug when @code{HANDLER} was used with some unsupported table type.
@item
@code{mysqldump} now puts @code{ALTER TABLE tbl_name DISABLE KEYS} and
@code{ALTER TABLE tbl_name ENABLE KEYS} in the sql dump.
@item
Added @code{mysql_fix_extensions} script.
@item
Fixed stack overrun problem with @code{LOAD DATA FROM MASTER} on OSF/1.
@item
Fixed shutdown problem on HP-UX.
@item
Added @code{DES_ENCRYPT()} and @code{DES_DECRYPT()} functions.
@item
Added @code{FLUSH DES_KEY_FILE} statement.
@item
Added @code{--des-key-file} option to @code{mysqld}.
@item
@code{HEX(string)} now returns the characters in @code{string} converted to
hexadecimal.
@item
Fixed problem with @code{GRANT} when using @code{lower_case_table_names == 1}.
@item
Changed @code{SELECT ... IN SHARE MODE} to
@code{SELECT ... LOCK IN SHARE MODE} (as in MySQL 3.23).
@item
A new query cache to cache results from identical @code{SELECT} queries.
@item
Fixed core dump bug on 64-bit machines when it got an incorrect communication
packet.
@item
@code{MATCH ... AGAINST(... IN BOOLEAN MODE)} can now work
without @code{FULLTEXT} index.
@item
Fixed slave to replicate from 3.23 master.
@item
Miscellaneous replication fixes/cleanup.
@item
Got shutdown to work on Mac OS X.
@item
Added @file{myisam/ft_dump} utility for low-level inspection
of @code{FULLTEXT} indexes.
@item
Fixed bug in @code{DELETE ... WHERE ... MATCH ...}.
@item
Added support for @code{MATCH ... AGAINST(... IN BOOLEAN MODE)}.
@strong{Note: you must rebuild your tables with
@code{ALTER TABLE tablename TYPE=MyISAM} to be
able to use boolean full-text search}.
@item
@code{LOCATE()} and @code{INSTR()} are now case-sensitive if either
argument is a binary string.
@item
Changed @code{RAND()} initialisation so that @code{RAND(N)} and
@code{RAND(N+1)} are more distinct.
@item
Fixed core dump bug in @code{UPDATE ... ORDER BY}.
@item
Changed @code{INSERT INTO ... SELECT} to stop on errors by default.
@item
Ignore @code{DATA DIRECTORY} and @code{INDEX DIRECTORY} directives on Windows.
@item
Added boolean full-text search code. It should be considered early alpha.
@item
Extended @code{MODIFY} and @code{CHANGE} in @code{ALTER TABLE} to accept
the @code{AFTER} keyword.
@item
Indexes are now used with @code{ORDER BY} on a whole @code{InnoDB} table.
@end itemize

@node News-4.0.0,  , News-4.0.1, News-4.0.x
@appendixsubsec Changes in release 4.0.0 (Oct 2001: Alpha)

@itemize @bullet
@item
Added @code{--xml} option to @code{mysql} for producing XML output.
@item
Added full-text variables @code{ft_min_word_len}, @code{ft_max_word_len}, and
@code{ft_max_word_len_for_sort}.
@item
Added documentation for @code{libmysqld}, the embedded MySQL server
library.  Also added example programs (a @code{mysql} client and
@code{mysqltest} test program) which use @code{libmysqld}.
@item
Removed all Gemini hooks from MySQL server.
@item
Removed @code{my_thread_init()} and @code{my_thread_end()}
from @file{mysql_com.h}, and added @code{mysql_thread_init()} and
@code{mysql_thread_end()} to @file{mysql.h}.
@item
Support for communication packets > 16M.  In 4.0.1 we will extend
@code{MyISAM} to be able to handle these.
@item
Secure connections (with SSL).
@item
Unsigned @code{BIGINT} constants now work. @code{MIN()} and @code{MAX()}
now handle signed and unsigned @code{BIGINT} numbers correctly.
@item
New character set @code{latin_de} which provides correct German sorting.
@item
@code{STRCMP()} now uses the current character set when doing comparisons,
which means that the default comparison behavior now is case-insensitive.
@item
@code{TRUNCATE TABLE} and @code{DELETE FROM tbl_name} are now separate
functions. One bonus is that @code{DELETE FROM tbl_name} now returns
the number of deleted rows, rather than zero.
@item
@code{DROP DATABASE} now executes a @code{DROP TABLE} on all tables in
the database, which fixes a problem with @code{InnoDB} tables.
@item
Added support for @code{UNION}.
@item
@code{DELETE} can now operate on multiple tables.
@item
A new @code{HANDLER} interface to @code{MyISAM} tables.
@item
Added support for @code{INSERT} on @code{MERGE} tables. Patch from
Benjamin Pflugmann.
@item
Changed @code{WEEK(#,0)} to match the calendar in the USA.
@item
@code{COUNT(DISTINCT)} is about 30% faster.
@item
Speed up all internal list handling.
@item
Speed up @code{IS NULL}, @code{ISNULL()} and some other internal primitives.
@item
Full-text index creation now is much faster.
@item
Tree-like cache to speed up bulk inserts and
@code{myisam_bulk_insert_tree_size} variable.
@item
Searching on packed (@code{CHAR}/@code{VARCHAR}) keys is now much faster.
@item
Optimised queries of type:
@code{SELECT DISTINCT * from tbl_name ORDER by key_part1 LIMIT #}.
@item
@code{SHOW CREATE TABLE} now shows all table attributes.
@item
@code{ORDER BY ... DESC} can now use keys.
@item
@code{LOAD DATA FROM MASTER} ``auto-magically'' sets up a slave.
@item
Renamed @code{safe_mysqld} to @code{mysqld_safe} to make this name more
in line with other MySQL scripts/commands.
@item
Added support for symbolic links to @code{MyISAM} tables. Symlink handling is
now enabled by default for Windows.
@item
Added @code{SQL_CALC_FOUND_ROWS} and @code{FOUND_ROWS()}. This makes it
possible to know how many rows a query would have returned
without a @code{LIMIT} clause.
@item
Changed output format of @code{SHOW OPEN TABLES}.
@item
Allow @code{SELECT expression LIMIT ...}.
@item
Added @code{IDENTITY} as a synonym for @code{AUTO_INCREMENT} (like Sybase).
@item
Added @code{ORDER BY} syntax to @code{UPDATE} and @code{DELETE}.
@item
@code{SHOW INDEXES} is now a synonym for @code{SHOW INDEX}.
@item
Added @code{ALTER TABLE tbl_name DISABLE KEYS} and
@code{ALTER TABLE tbl_name ENABLE KEYS} commands.
@item
Allow use of @code{IN} as a synonym for @code{FROM} in @code{SHOW} commands.
@item
Implemented ``repair by sort'' for @code{FULLTEXT} indexes.
@code{REPAIR TABLE}, @code{ALTER TABLE}, and @code{OPTIMIZE TABLE}
for tables with @code{FULLTEXT} indexes are now up to 100 times faster.
@item
Allow ANSI SQL syntax @code{X'hexadecimal-number'}.
@item
Cleaned up global lock handling for @code{FLUSH TABLES WITH READ LOCK}.
@item
Fixed problem with @code{DATETIME = constant} in @code{WHERE} optimisation.
@item
Added @code{--master-data} and @code{--no-autocommit} options to
@code{mysqldump}. (Thanks to Brian Aker for this.)
@item
Added script @code{mysql_explain_log.sh} to distribution.
(Thanks to mobile.de).
@end itemize


@node News-3.23.x, News-3.22.x, News-4.0.x, News
@appendixsec Changes in release 3.23.x  (Stable)

@cindex changes, version 3.23

The 3.23 release has several major features that are not
present in previous versions.  We have added three new
table types:

@table @asis
@item @code{MyISAM}
A new ISAM library which is tuned for SQL and supports large files.
@item @code{BerkeleyDB} or @code{BDB}
Uses the Berkeley DB library from Sleepycat Software to implement
transaction-safe tables.
@item @code{InnoDB}
A transaction-safe table handler that supports row level locking, and many
Oracle-like features.
@end table

Note that only @code{MyISAM} is available in the standard binary distribution.

The 3.23 release also includes support for database replication between
a master and many slaves, full-text indexing, and much more.

All new features are being developed in the 4.0 version.  Only
bug fixes and minor enhancements to existing features will be
added to 3.23.

The replication code and BerkeleyDB code is still not as tested and as
the rest of the code, so we will probably need to do a couple of future
releases of 3.23 with small fixes for this part of the code. As long as
you don't use these features, you should be quite safe with
MySQL 3.23!

Note that the above doesn't mean that replication or Berkeley DB don't
work.  We have done a lot of testing of all code, including replication
and BDB without finding any problems.  It only means that not as many
users use this code as the rest of the code and because of this we are
not yet 100% confident in this code.

@menu
* News-3.23.51::                Changes in release 3.23.51
* News-3.23.50::                Changes in release 3.23.50 (21 Apr 2002)
* News-3.23.49::                Changes in release 3.23.49
* News-3.23.48::                Changes in release 3.23.48 (07 Feb 2002)
* News-3.23.47::                Changes in release 3.23.47 (27 Dec 2001)
* News-3.23.46::                Changes in release 3.23.46 (29 Nov 2001)
* News-3.23.45::                Changes in release 3.23.45 (22 Nov 2001)
* News-3.23.44::                Changes in release 3.23.44 (31 Oct 2001)
* News-3.23.43::                Changes in release 3.23.43
* News-3.23.42::                Changes in release 3.23.42 (08 Sep 2001)
* News-3.23.41::                Changes in release 3.23.41 (11 Aug 2001)
* News-3.23.40::                Changes in release 3.23.40
* News-3.23.39::                Changes in release 3.23.39 (12 Jun 2001)
* News-3.23.38::                Changes in release 3.23.38 (09 May 2001)
* News-3.23.37::                Changes in release 3.23.37 (17 Apr 2001)
* News-3.23.36::                Changes in release 3.23.36 (27 Mar 2001)
* News-3.23.35::                Changes in release 3.23.35 (15 Mar 2001)
* News-3.23.34a::               Changes in release 3.23.34a
* News-3.23.34::                Changes in release 3.23.34 (10 Mar 2001)
* News-3.23.33::                Changes in release 3.23.33 (09 Feb 2001)
* News-3.23.32::                Changes in release 3.23.32 (22 Jan 2001: Stable)
* News-3.23.31::                Changes in release 3.23.31 (17 Jan 2001)
* News-3.23.30::                Changes in release 3.23.30 (04 Jan 2001)
* News-3.23.29::                Changes in release 3.23.29 (16 Dec 2000)
* News-3.23.28::                Changes in release 3.23.28 (22 Nov 2000: Gamma)
* News-3.23.27::                Changes in release 3.23.27 (24 Oct 2000)
* News-3.23.26::                Changes in release 3.23.26
* News-3.23.25::                Changes in release 3.23.25
* News-3.23.24::                Changes in release 3.23.24 (08 Sep 2000)
* News-3.23.23::                Changes in release 3.23.23
* News-3.23.22::                Changes in release 3.23.22 (31 Jul 2000)
* News-3.23.21::                Changes in release 3.23.21
* News-3.23.20::                Changes in release 3.23.20
* News-3.23.19::                Changes in release 3.23.19
* News-3.23.18::                Changes in release 3.23.18
* News-3.23.17::                Changes in release 3.23.17
* News-3.23.16::                Changes in release 3.23.16
* News-3.23.15::                Changes in release 3.23.15 (May 2000: Beta)
* News-3.23.14::                Changes in release 3.23.14
* News-3.23.13::                Changes in release 3.23.13
* News-3.23.12::                Changes in release 3.23.12
* News-3.23.11::                Changes in release 3.23.11
* News-3.23.10::                Changes in release 3.23.10
* News-3.23.9::                 Changes in release 3.23.9
* News-3.23.8::                 Changes in release 3.23.8
* News-3.23.7::                 Changes in release 3.23.7
* News-3.23.6::                 Changes in release 3.23.6
* News-3.23.5::                 Changes in release 3.23.5
* News-3.23.4::                 Changes in release 3.23.4
* News-3.23.3::                 Changes in release 3.23.3
* News-3.23.2::                 Changes in release 3.23.2
* News-3.23.1::                 Changes in release 3.23.1
* News-3.23.0::                 Changes in release 3.23.0 (Sep 1999: Alpha)
@end menu

@node News-3.23.51, News-3.23.50, News-3.23.x, News-3.23.x
@appendixsubsec Changes in release 3.23.51

@itemize @bullet
@item
Remove end space from @code{enum} values. (This fixed a problem with
@code{SHOW CREATE TABLE}).
@item
Fixed bug in @code{CONCAT_WS()} that cut the result.
@item
Changed name of variables @code{Com_show_master_stat} to
@code{Com_show_master_status} and @code{Com_show_slave_stat} to
@code{Com_show_slave_status}.
@item
Changed handling of @code{gethostbyname()} to make the client library
threadsafe even if @code{gethostbyname_r} doesn't exists.
@item
Fixed core-dump problem when giving a wrong password string to @code{GRANT}.
@item
Fixed bug in @code{DROP DATABASE} with symlinked directory.
@item
Fixed optimization problem with @code{DATETIME} and value outside
@code{DATETIME} range.
@item
Removed BDB documentation.
@item
Fixed mit-pthreads to compile with glibc 2.2 (needed for @code{make dist}).
@item
Fixed the @code{FLOAT(X+1,X)} is not converted to @code{FLOAT(X+2,X)}.
(This also affected @code{DECIMAL}, @code{DOUBLE} and @code{REAL} types)
@item
Fixed the result from @code{IF()} is case in-sensitive if the 2 and
third arguments are case sensitive.
@item
Fixed core dump problem on OSF in @code{gethostbyname_r}.
@item
Fixed that underflowed decimal fields is not zero filled.
@item
If we get an overflow when inserting @code{'+11111'} for
@code{decimal(5,0) unsigned} columns, we will just drop the sign.
@item
Fixed optimization bug with @code{ISNULL(expression_which_cannot_be_null)} and
@code{ISNULL(constant_expression)}.
@item
Fixed host lookup bug in the glibc library that we used with the 3.23.50
Linux-x86 binaries.
@end itemize

@node News-3.23.50, News-3.23.49, News-3.23.51, News-3.23.x
@appendixsubsec Changes in release 3.23.50 (21 Apr 2002)

@itemize @bullet
@item
Fixed problem with @code{crash-me} and @code{gcc} 3.0.4.
@item
Fixed that @code{@@@@unknown_variable} doesn't hang server.
@item
Added @code{@@@@VERSION} as a synonym for @code{VERSION()}.
@item
@code{SHOW VARIABLES LIKE 'xxx'} is now case-insensitive.
@item
Fixed timeout for @code{GET_LOCK()} on HPUX with DCE threads.
@item
Fixed memory allocation bug in the glibc library used to build Linux
binaries, which caused mysqld to die in 'free()'.
@item
Fixed @code{SIGINT} and @code{SIGQUIT} problems in @code{mysql}.
@item
Fixed bug in character table converts when used with big ( > 64K) strings.
@item
@code{InnoDB} now retains foreign key constraints through @code{ALTER TABLE}
and @code{CREATE/DROP INDEX}.
@item
@code{InnoDB} now allows foreign key constraints to be added through the
@code{ALTER TABLE} syntax.
@item
@code{InnoDB} tables can now be set to automatically grow in size (autoextend).
@item
Our Linux RPMS and binaries are now compiled with @code{gcc} 3.0.4, which
should make them a bit faster.
@item
Fixed some buffer overflow problems when reading startup parameters.
@item
Because of problems on shutdown we have now disabled named pipes on
windows by default.  One can enable named pipes by starting mysqld with
@code{--enable-named-pipe}.
@item
Fixed bug when using @code{WHERE key_column = 'J' or key_column='j'}.
@item
Fixed core-dump bug when using @code{--log-bin} with @code{LOAD DATA
INFILE} without an active database.
@item
Fixed bug in @code{RENAME TABLE} when used with
@code{lower_case_table_names=1} (default on Windows).
@item
Fixed unlikely core-dump bug when using @code{DROP TABLE} on a table
that was in use by a thread that also used queries on only temporary tables.
@item
Fixed problem with @code{SHOW CREATE TABLE} and @code{PRIMARY KEY} when using
32 indexes.
@item
Fixed that one can use @code{SET PASSWORD} for the anonymous user.
@item
Fixed core dump bug when reading client groups from option files using
@code{mysql_options()}.
@item
Memory leak (16 bytes per every @strong{corrupted} table) closed.
@item
Fixed binary builds to use @code{--enable-local-infile}.
@item
Update source to work with new version of @code{bison}.
@item
Updated shell scripts to now agree with new POSIX standard.
@item
Fixed bug where @code{DATE_FORMAT()} returned empty string when used
with @code{GROUP BY}.
@end itemize

@node News-3.23.49, News-3.23.48, News-3.23.50, News-3.23.x
@appendixsubsec Changes in release 3.23.49

@itemize @bullet
@item
Don't give warning for a statement that is only a comment; this is needed for
@code{mysqldump --disable-keys} to work.
@item
Fixed unlikely caching bug when doing a join without keys. In this case
the last used field for a table always returned @code{NULL}.
@item
Added options to make @code{LOAD DATA LOCAL INFILE} more secure.
@item
MySQL binary release 3.23.48 for Linux contained a new @code{glibc} library,
which has serious problems under high load and RedHat 7.2.  The 3.23.49 binary
release doesn't have this problem.
@item
Fixed shutdown problem on NT.
@end itemize

@node News-3.23.48, News-3.23.47, News-3.23.49, News-3.23.x
@appendixsubsec Changes in release 3.23.48 (07 Feb 2002)

@itemize @bullet
@item
Added @code{--xml} option to @code{mysqldump} for producing XML output.
@item
Changed to use @code{autoconf} 2.52 (from @code{autoconf} 2.13)
@item
Fixed bug in complicated join with @code{const} tables.
@item
Added internal safety checks for @code{InnoDB}.
@item
Some @code{InnoDB} variables were always shown in @code{SHOW VARIABLES} as
@code{OFF} on high-byte-first systems (like SPARC).
@item
Fixed problem with one thread using an @code{InnoDB} table and another
thread doing an @code{ALTER TABLE} on the same table. Before that,
@code{mysqld} could crash with an assertion failure in @file{row0row.c},
line 474.
@item
Tuned the @code{InnoDB} SQL optimiser to favor index searches more often
over table scans.
@item
Fixed a performance problem with @code{InnoDB} tables when several large
@code{SELECT} queries are run concurrently on a multiprocessor Linux
computer.  Large CPU-bound @code{SELECT} queries will now also generally
run faster on all platforms.
@item
If MySQL binlogging is used, @code{InnoDB} now prints after crash recovery the
latest MySQL binlog name and the offset @code{InnoDB} was able to recover
to. This is useful, for example, when resynchronising a master and a
slave database in replication.
@item
Added better error messages to help in installation problems of @code{InnoDB}
tables.
@item
It is now possible to recover MySQL temporary tables that have become
orphaned inside the @code{InnoDB} tablespace.
@item
@code{InnoDB} now prevents a @code{FOREIGN KEY} declaration where the
signedness is not the same in the referencing and referenced integer columns.
@item
Calling @code{SHOW CREATE TABLE} or @code{SHOW TABLE STATUS} could cause
memory corruption and make @code{mysqld} crash.  Especially at risk was
@code{mysqldump}, because it frequently calls @code{SHOW CREATE TABLE}.
@item
If inserts to several tables containing an @code{AUTO_INCREMENT} column were
wrapped inside one @code{LOCK TABLES}, @code{InnoDB} asserted in
@file{lock0lock.c}.
@item
In 3.23.47 we allowed several @code{NULL} values in a @code{UNIQUE} secondary
index for an @code{InnoDB} table.  But @code{CHECK TABLE} was not relaxed: it
reports the table as corrupt.  @code{CHECK TABLE} no longer complains in
this situation.
@item
@code{SHOW GRANTS} now shows @code{REFERENCES} instead of @code{REFERENCE}.
@end itemize

@node News-3.23.47, News-3.23.46, News-3.23.48, News-3.23.x
@appendixsubsec Changes in release 3.23.47 (27 Dec 2001)

@itemize @bullet
@item
Fixed bug when using the following construct:
@code{SELECT ... WHERE key=@@var_name OR key=@@var_name2}
@item
Restrict @code{InnoDB} keys to 500 bytes.
@item
@code{InnoDB} now supports @code{NULL} in keys.
@item
Fixed shutdown problem on HP-UX. (Introduced in 3.23.46)
@item
Fixed core dump bug in replication when using @code{SELECT RELEASE_LOCK()}.
@item
Added new command: @code{DO expression,[expression]}
@item
Added @code{slave-skip-errors} option.
@item
Added statistics variables for all MySQL commands. (@code{SHOW STATUS} is
now much longer.)
@item
Fixed default values for @code{InnoDB} tables.
@item
Fixed that @code{GROUP BY expr DESC} works.
@item
Fixed bug when using @code{t1 LEFT JOIN t2 ON t2.key=constant}.
@item
@code{mysql_config} now also works with binary (relocated) distributions.
@end itemize

@node News-3.23.46, News-3.23.45, News-3.23.47, News-3.23.x
@appendixsubsec Changes in release 3.23.46 (29 Nov 2001)

@itemize @bullet
@item
Fixed problem with aliased temporary table replication.
@item
@code{InnoDB} and @code{BDB} tables will now use index when doing an
@code{ORDER BY} on the whole table.
@item
Fixed bug where one got an empty set instead of a DEADLOCK error when using
@code{BDB} tables.
@item
One can now kill @code{ANALYZE}, @code{REPAIR}, and @code{OPTIMIZE TABLE} when
the thread is waiting to get a lock on the table.
@item
Fixed race condition in @code{ANALYZE TABLE}.
@item
Fixed bug when joining with caching (unlikely to happen).
@item
Fixed race condition when using the binary log and @code{INSERT DELAYED}
which could cause the binary log to have rows that were not yet written
to @code{MyISAM} tables.
@item
Changed caching of binary log to make replication slightly faster.
@item
Fixed bug in replication on Mac OS X.
@end itemize

@node News-3.23.45, News-3.23.44, News-3.23.46, News-3.23.x
@appendixsubsec Changes in release 3.23.45 (22 Nov 2001)

@itemize @bullet
@item
@code{(UPDATE|DELETE) ...WHERE MATCH} bugfix.
@item
shutdown should now work on Darwin (Mac OS X).
@item
Fixed core dump when repairing corrupted packed @code{MyISAM} files.
@item
@code{--core-file} now works on Solaris.
@item
Fix a bug which could cause @code{InnoDB} to complain if it cannot find
free blocks from the buffer cache during recovery.
@item
Fixed bug in @code{InnoDB} insert buffer B-tree handling that could cause
crashes.
@item
Fixed bug in @code{InnoDB} lock timeout handling.
@item
Fixed core dump bug in @code{ALTER TABLE} on a @code{TEMPORARY} @code{InnoDB}
table.
@item
Fixed bug in @code{OPTIMIZE TABLE} that reset index cardinality if it
was up to date.
@item
Fixed problem with @code{t1 LEFT_JOIN t2 ... WHERE t2.date_column IS NULL} when
date_column was declared as @code{NOT NULL}.
@item
Fixed bug with @code{BDB} tables and keys on @code{BLOB} columns.
@item
Fixed bug in @code{MERGE} tables on OS with 32-bit file pointers.
@item
Fixed bug in @code{TIME_TO_SEC()} when using negative values.
@end itemize

@node News-3.23.44, News-3.23.43, News-3.23.45, News-3.23.x
@appendixsubsec Changes in release 3.23.44 (31 Oct 2001)

@itemize @bullet
@item
Fixed @code{Rows_examined} count in slow query log.
@item
Fixed bug when using a reference to an @code{AVG()} column in @code{HAVING}.
@item
Fixed that date functions that require correct dates, like
@code{DAYOFYEAR(column)}, will return @code{NULL} for @code{0000-00-00} dates.
@item
Fixed bug in const-propagation when comparing columns of different
types. (@code{SELECT * FROM date_col="2001-01-01" and date_col=time_col})
@item
Fixed bug that caused error message @code{Can't write, because of unique
constraint} with some @code{GROUP BY} queries.
@item
Fixed problem with @code{sjis} character strings used within quoted table
names.
@item
Fixed core dump when using @code{CREATE ... FULLTEXT} keys with other table
handlers than @code{MyISAM}.
@item
Don't use @code{signal()} on Windows because this appears to not be
100 % reliable.
@item
Fixed bug when doing @code{WHERE col_name=NULL} on an indexed column
that had @code{NULL} values.
@item
Fixed bug when doing @code{LEFT JOIN ...  ON (col_name = constant) WHERE col_name = constant}.
@item
When using replications, aborted queries that contained @code{%} could cause
a core dump.
@item
@code{TCP_NODELAY} was not used on some systems. (Speed problem.)
@item
Applied portability fixes for OS/2. (Patch by Yuri Dario.)
@end itemize

The following changes are for @code{InnoDB} tables:

@itemize @bullet
@item
Add missing @code{InnoDB} variables to @code{SHOW VARIABLES}.
@item
Foreign keys checking is now done for @code{InnoDB} tables.
@item
@code{DROP DATABASE} now works also for @code{InnoDB} tables.
@item
@code{InnoDB} now supports datafiles and raw disk partitions bigger
than 4 GB on those operating systems that have big files.
@item
@code{InnoDB} calculates better table cardinality estimates for the
MySQL optimiser.
@item
Accent characters in the default character set @code{latin1} are ordered
according to the MySQL ordering.

Note: if you are using @code{latin1} and have inserted characters whose
code is greater than 127 into an indexed @code{CHAR} column, you should
run @code{CHECK TABLE} on your table when you upgrade to 3.23.44, and
drop and reimport the table if @code{CHECK TABLE} reports an error!
@item
A new @file{my.cnf} parameter, @code{innodb_thread_concurrency}, helps in
performance tuning in heavily concurrent environments.
@item
A new @file{my.cnf} parameter, @code{innodb_fast_shutdown}, speeds up
server shutdown.
@item
A new @file{my.cnf} parameter, @code{innodb_force_recovery}, helps to save
your data in case the disk image of the database becomes corrupt.
@item
@code{innodb_monitor} has been improved and a new
@code{innodb_table_monitor} added.
@item
Increased maximum key length from 500 to 7000 bytes.
@item
Fixed a bug in replication of @code{AUTO_INCREMENT} columns with
multiple-line inserts.
@item
Fixed a bug when the case of letters changes in an update of an indexed
secondary column.
@item
Fixed a hang when there are > 24 datafiles.
@item
Fixed a crash when @code{MAX(col)} is selected from an empty table, and
@code{col} is not the first column in a multi-column index.
@item
Fixed a bug in purge which could cause crashes.
@end itemize

@node News-3.23.43, News-3.23.42, News-3.23.44, News-3.23.x
@appendixsubsec Changes in release 3.23.43

@itemize @bullet
@item
Fixed a bug in @code{INSERT DELAYED} and @code{FLUSH TABLES} introduced
in 3.23.42.
@item
Fixed unlikely bug, which returned non-matching rows, in @code{SELECT} with
many tables and multi-column indexes and 'range' type.
@item
Fixed an unlikely core dump bug when doing @code{EXPLAIN SELECT} when using
many tables and @code{ORDER BY}.
@item
Fixed bug in @code{LOAD DATA FROM MASTER} when using table with
@code{CHECKSUM=1}.
@item
Added unique error message when one gets a DEADLOCK during a transaction with
@code{BDB} tables.
@item
Fixed problem with @code{BDB} tables and @code{UNIQUE} columns defined
as @code{NULL}.
@item
Fixed problem with @code{myisampack} when using pre-space filled @code{CHAR}
columns.
@item
Applied patch from Yuri Dario for OS/2.
@item
Fixed bug in @code{--safe-user-create}.
@end itemize

@node News-3.23.42, News-3.23.41, News-3.23.43, News-3.23.x
@appendixsubsec Changes in release 3.23.42 (08 Sep 2001)

@itemize @bullet
@item
Fixed problem when using @code{LOCK TABLES} and @code{BDB} tables.
@item
Fixed problem with @code{REPAIR TABLE} on @code{MyISAM} tables with row
lengths in the range from 65517 to 65520 bytes.
@item
Fixed rare hang when doing @code{mysqladmin shutdown} when there was
a lot of activity in other threads.
@item
Fixed problem with @code{INSERT DELAYED} where delay thread could be
hanging on @code{upgrading locks} with no apparent reason.
@item
Fixed problem with @code{myisampack} and @code{BLOB}.
@item
Fixed problem when one edited @file{.MRG} tables by hand.
(Patch from Benjamin Pflugmann).
@item
Enforce that all tables in a @code{MERGE} table come from the same
database.
@item
Fixed bug with @code{LOAD DATA INFILE} and transactional tables.
@item
Fix bug when using @code{INSERT DELAYED} with wrong column definition.
@item
Fixed core dump during @code{REPAIR} of some particularly broken tables.
@item
Fixed bug in @code{InnoDB} and @code{AUTO_INCREMENT} columns.
@item
Fixed bug in @code{InnoDB} and @code{RENAME TABLE} columns.
@item
Fixed critical bug in @code{InnoDB} and @code{BLOB} columns.  If you have
used @code{BLOB} columns larger than 8000 bytes in an @code{InnoDB}
table, it is necessary to dump the table with @code{mysqldump}, drop it and
restore it from the dump.
@item
Applied large patch for OS/2 from Yuri Dario.
@item
Fixed problem with @code{InnoDB} when one could get the error @code{Can't
execute the given command...} even when no transaction was active.
@item
Applied some minor fixes that concern Gemini.
@item
Use real arithmetic operations even in integer context if not
all arguments are integers. (Fixes uncommon bug in some integer
contexts).
@item
Don't force everything to lowercase on Windows. (To fix problem
with Windows and @code{ALTER TABLE}).  Now @code{--lower_case_names}
also works on Unix.
@item
Fixed that automatic rollback is done when thread end doesn't lock
other threads.
@end itemize

@node News-3.23.41, News-3.23.40, News-3.23.42, News-3.23.x
@appendixsubsec Changes in release 3.23.41 (11 Aug 2001)

@itemize @bullet
@item
Added @code{--sql-mode=option[,option[,option]]} option to @code{mysqld}.
@xref{Command-line options}.
@item
Fixed possible problem with @code{shutdown} on Solaris where the
@file{.pid} file wasn't deleted.
@item
@code{InnoDB} now supports < 4 GB rows. The former limit was 8000 bytes.
@item
The @code{doublewrite} file flush method is used in @code{InnoDB}.
It reduces the need for Unix @code{fsync()} calls to a fraction and
improves performance on most Unix flavors.
@item
You can now use the @code{InnoDB} Monitor to print a lot of @code{InnoDB} state
information, including locks, to the standard output.  This is useful in
performance tuning.
@item
Several bugs which could cause hangs in @code{InnoDB} have been fixed.
@item
Split @code{record_buffer} to @code{record_buffer} and
@code{record_rnd_buffer}.  To make things compatible to previous MySQL
versions, if @code{record_rnd_buffer} is not set, then it takes the
value of @code{record_buffer}.
@item
Fixed optimising bug in @code{ORDER BY} where some @code{ORDER BY} parts
where wrongly removed.
@item
Fixed overflow bug with @code{ALTER TABLE} and @code{MERGE} tables.
@item
Added prototypes for @code{my_thread_init()} and @code{my_thread_end()} to
@file{mysql_com.h}
@item
Added @code{--safe-user-create} option to @code{mysqld}.
@item
Fixed bug in @code{SELECT DISTINCT ... HAVING} that caused error message
@code{Can't find record in #...}
@end itemize


@node News-3.23.40, News-3.23.39, News-3.23.41, News-3.23.x
@appendixsubsec Changes in release 3.23.40

@itemize @bullet
@item
Fixed problem with @code{--low-priority-updates} and @code{INSERT} statements.
@item
Fixed bug in slave thread when under some rare circumstances it could
get 22 bytes ahead on the offset in the master.
@item
Added @code{slave_net_timeout} for replication.
@item
Fixed problem with @code{UPDATE} and @code{BDB} tables.
@item
Fixed hard bug in @code{BDB} tables when using key parts.
@item
Fixed problem when using @code{GRANT FILE ON database.* ...}; previously
we added the @strong{drop} privilege for the database.
@item
Fixed @code{DELETE FROM tbl_name ... LIMIT 0} and
@code{UPDATE FROM tbl_name ... LIMIT 0}, which acted as though the
@code{LIMIT} clause was not present (they deleted or updated all selected
rows).
@item
@code{CHECK TABLE} now checks if an @code{AUTO_INCREMENT} column contains
the value 0.
@item
Sending a @code{SIGHUP} to @code{mysqld} will now only flush the logs,
not reset the replication.
@item
Fixed parser to allow floats of type @code{1.0e1} (no sign after @code{e}).
@item
Option @code{--force} to @code{myisamchk} now also updates states.
@item
Added option @code{--warnings} to @code{mysqld}. Now @code{mysqld}
prints the error @code{Aborted connection} only if this option is used.
@item
Fixed problem with @code{SHOW CREATE TABLE} when you didn't have a
@code{PRIMARY KEY}.
@item
Properly fixed the rename of @code{innodb_unix_file_flush_method} variable to
@code{innodb_flush_method}.
@item
Fixed bug when converting @code{BIGINT UNSIGNED} to @code{DOUBLE}. This caused
a problem when doing comparisons with @code{BIGINT} values outside of the
signed range.
@item
Fixed bug in @code{BDB} tables when querying empty tables.
@item
Fixed a bug when using @code{COUNT(DISTINCT)} with @code{LEFT JOIN} and
there weren't any matching rows.
@item
Removed all documentation referring to the @code{GEMINI} table
type. @code{GEMINI} is not released under an @code{Open Source} license.
@end itemize


@node News-3.23.39, News-3.23.38, News-3.23.40, News-3.23.x
@appendixsubsec Changes in release 3.23.39 (12 Jun 2001)

@itemize @bullet
@item
The @code{AUTO_INCREMENT} sequence wasn't reset when dropping
and adding an @code{AUTO_INCREMENT} column.
@item
@code{CREATE ... SELECT} now creates non-unique indexes delayed.
@item
Fixed problem where @code{LOCK TABLES tbl_name READ} followed by
@code{FLUSH TABLES} put an exclusive lock on the table.
@item
@code{REAL @@variable} values were represented with only 2 digits when
converted to strings.
@item
Fixed problem that client ``hung'' when @code{LOAD TABLE FROM MASTER} failed.
@item
@code{myisamchk --fast --force} will no longer repair tables
that only had the open count wrong.
@item
Added functions to handle symbolic links to make life easier in 4.0.
@item
We are now using the @code{-lcma} thread library on HP-UX 10.20 so
that MySQL will be more stable on HP-UX.
@item
Fixed problem with @code{IF()} and number of decimals in the result.
@item
Fixed date-part extraction functions to work with dates where day
and/or month is 0.
@item
Extended argument length in option files from 256 to 512 chars.
@item
Fixed problem with shutdown when @code{INSERT DELAYED} was waiting for
a @code{LOCK TABLE}.
@item
Fixed core dump bug in @code{InnoDB} when tablespace was full.
@item
Fixed problem with @code{MERGE} tables and big tables (> 4G) when using
@code{ORDER BY}.
@end itemize


@node News-3.23.38, News-3.23.37, News-3.23.39, News-3.23.x
@appendixsubsec Changes in release 3.23.38 (09 May 2001)

@itemize @bullet
@item
Fixed a bug when @code{SELECT} from @code{MERGE} table
sometimes results in incorrectly ordered rows.
@item
Fixed a bug in @code{REPLACE()} when using the @code{ujis} character set.
@item
Applied Sleepycat @code{BDB} patches 3.2.9.1 and 3.2.9.2.
@item
Added @code{--skip-stack-trace} option to @code{mysqld}.
@item
@code{CREATE TEMPORARY} now works with @code{InnoDB} tables.
@item
@code{InnoDB} now promotes sub keys to whole keys.
@item
Added option @code{CONCURRENT} to @code{LOAD DATA}.
@item
Better error message when slave @code{max_allowed_packet} is too low to
read a very long log event from the master.
@item
Fixed bug when too many rows where removed when using
@code{SELECT DISTINCT ... HAVING}.
@item
@code{SHOW CREATE TABLE} now returns @code{TEMPORARY} for temporary tables.
@item
Added @code{Rows_examined} to slow query log.
@item
Fixed problems with function returning empty string when used together with
a group function and a @code{WHERE} that didn't match any rows.
@item
New program @code{mysqlcheck}.
@item
Added database name to output for administrative commands like @code{CHECK},
@code{REPAIR}, @code{OPTIMIZE}.
@item
Lots of portability fixes for @code{InnoDB}.
@item
Changed optimiser so that queries like
@code{SELECT * FROM tbl_name,tbl_name2 ... ORDER BY key_part1 LIMIT #}
will use index on @code{key_part1} instead of @code{filesort}.
@item
Fixed bug when doing
@code{LOCK TABLE to_table WRITE,...; INSERT INTO to_table... SELECT ...}
when @code{to_table} was empty.
@item
Fixed bug with @code{LOCK TABLE} and @code{BDB} tables.
@end itemize


@node News-3.23.37, News-3.23.36, News-3.23.38, News-3.23.x
@appendixsubsec Changes in release 3.23.37 (17 Apr 2001)

@itemize @bullet
@item
Fixed a bug when using @code{MATCH()} in @code{HAVING} clause.
@item
Fixed a bug when using @code{HEAP} tables with @code{LIKE}.
@item
Added @code{--mysql-version} option to @code{safe_mysqld}
@item
Changed @code{INNOBASE} to @code{InnoDB} (because the @code{INNOBASE}
name was already used). All @code{configure} options and @code{mysqld}
start options now use @code{innodb} instead of @code{innobase}. This
means that before upgrading to this version, you have to change any
configuration files where you have used @code{innobase} options!
@item
Fixed bug when using indexes on @code{CHAR(255) NULL} columns.
@item
Slave thread will now be started even if @code{master-host} is not set, as
long as @code{server-id} is set and valid @file{master.info} is present.
@item
Partial updates (terminated with kill) are now logged with a special error
code to the binary log. Slave will refuse to execute them if the error code
indicates the update was terminated abnormally, and will have to be recovered
with @code{SET SQL_SLAVE_SKIP_COUNTER=1; SLAVE START} after a manual sanity
check/correction of data integrity.
@item
Fixed bug that erroneously logged a drop of internal temporary table
on thread termination to the binary log -- this bug affected replication.
@item
Fixed a bug in @code{REGEXP} on 64-bit machines.
@item
@code{UPDATE} and @code{DELETE} with @code{WHERE unique_key_part IS NULL}
didn't update/delete all rows.
@item
Disabled @code{INSERT DELAYED} for tables that support transactions.
@item
Fixed bug when using date functions on @code{TEXT}/@code{BLOB} column
with wrong date format.
@item
UDFs now also work on Windows. (Patch by Ralph Mason.)
@item
Fixed bug in @code{ALTER TABLE} and @code{LOAD DATA INFILE} that disabled
key-sorting. These commands should now be faster in most cases.
@item
Fixed performance bug where reopened tables (tables that had been
waiting for @code{FLUSH} or @code{REPAIR}) would not use indexes for the
next query.
@item
Fixed problem with @code{ALTER TABLE} to @code{InnoDB} tables on FreeBSD.
@item
Added @code{mysqld} variables @code{myisam_max_sort_file_size} and
@code{myisam_max_extra_sort_file_size}.
@item
Initialise signals early to avoid problem with signals in @code{InnoDB}.
@item
Applied patch for the @code{tis620} character set to make comparisons
case-independent and to fix a bug in @code{LIKE} for this character set.
@strong{Note}: All tables that uses the @code{tis620} character set must be
fixed with  @code{myisamchk -r} or @code{REPAIR TABLE} !
@item
Added @code{--skip-safemalloc} option to @code{mysqld}.
@end itemize


@node News-3.23.36, News-3.23.35, News-3.23.37, News-3.23.x
@appendixsubsec Changes in release 3.23.36 (27 Mar 2001)

@itemize @bullet
@item
Fixed a bug that allowed use of database names containing a @samp{.}
character.  This fixes a serious security issue when @code{mysqld} is run
as root.
@item
Fixed bug when thread creation failed (could happen when doing a @strong{lot}
of connections in a short time).
@item
Fixed some problems with @code{FLUSH TABLES} and @code{TEMPORARY} tables.
(Problem with freeing the key cache and error @code{Can't reopen table...}.)
@item
Fixed a problem in @code{InnoDB} with other character sets than @code{latin1}
and another problem when using many columns.
@item
Fixed bug that caused a core dump when using a very complex query involving
@code{DISTINCT} and summary functions.
@item
Added @code{SET TRANSACTION ISOLATION LEVEL ...}
@item
Added @code{SELECT ... FOR UPDATE}.
@item
Fixed bug where the number of affected rows was not returned when MySQL
was compiled without transaction support.
@item
Fixed a bug in @code{UPDATE} where keys weren't always used to find the
rows to be updated.
@item
Fixed a bug in @code{CONCAT_WS()} where it returned incorrect results.
@item
Changed @code{CREATE ... INSERT} and @code{INSERT ... SELECT} to not
allow concurrent inserts as this could make the binary log hard to repeat.
(Concurrent inserts are enabled if you are not using the binary or update log.)
@item
Changed some macros to be able to use fast mutex with @code{glibc} 2.2.
@end itemize


@node News-3.23.35, News-3.23.34a, News-3.23.36, News-3.23.x
@appendixsubsec Changes in release 3.23.35 (15 Mar 2001)

@itemize @bullet
@item
Fixed newly introduced bug in @code{ORDER BY}.
@item
Fixed wrong define @code{CLIENT_TRANSACTIONS}.
@item
Fixed bug in @code{SHOW VARIABLES} when using @code{INNOBASE} tables.
@item
Setting and using user variables in @code{SELECT DISTINCT} didn't work.
@item
Tuned @code{SHOW ANALYZE} for small tables.
@item
Fixed handling of arguments in the benchmark script @code{run-all-tests}.
@end itemize


@node News-3.23.34a, News-3.23.34, News-3.23.35, News-3.23.x
@appendixsubsec Changes in release 3.23.34a

@itemize @bullet
@item
Added extra files to the distribution to allow @code{INNOBASE} support
to be compiled.
@end itemize


@node News-3.23.34, News-3.23.33, News-3.23.34a, News-3.23.x
@appendixsubsec Changes in release 3.23.34 (10 Mar 2001)

@itemize @bullet
@item
Added the @code{INNOBASE} table handler and the @code{BDB} table handler
to the MySQL source distribution.
@item
Updated the documentation about @code{GEMINI} tables.
@item
Fixed a bug in @code{INSERT DELAYED} that caused threads to hang when
inserting @code{NULL} into an @code{AUTO_INCREMENT} column.
@item
Fixed a bug in @code{CHECK TABLE} / @code{REPAIR TABLE} that could cause
a thread to hang.
@item
@code{REPLACE} will not replace a row that conflicts with an
@code{AUTO_INCREMENT} generated key.
@item
@code{mysqld} now only sets @code{CLIENT_TRANSACTIONS} in
@code{mysql->server_capabilities} if the server supports a
transaction-safe handler.
@item
Fixed @code{LOAD DATA INFILE} to allow numeric values to be read into
@code{ENUM} and @code{SET} columns.
@item
Improved error diagnostic for slave thread exit.
@item
Fixed bug in @code{ALTER TABLE ... ORDER BY}.
@item
Added @code{max_user_connections} variable to @code{mysqld}.
@item
Limit query length for replication by @code{max_allowed_packet}, not the
arbitrary limit of 4 MB.
@item
Allow space around @code{=} in argument to @code{--set-variable}.
@item
Fixed problem in automatic repair that could leave some threads in state
@code{Waiting for table}.
@item
@code{SHOW CREATE TABLE} now displays the @code{UNION()} for @code{MERGE}
tables.
@item
@code{ALTER TABLE} now remembers the old @code{UNION()} definition.
@item
Fixed bug when replicating timestamps.
@item
Fixed bug in bidirectional replication.
@item
Fixed bug in the @code{BDB} table handler that occurred when using an index
on multi-part key where a key part may be @code{NULL}.
@item
Fixed @code{MAX()} optimisation on sub-key for @code{BDB} tables.
@item
Fixed problem where garbage results were returned when using @code{BDB}
tables and @code{BLOB} or @code{TEXT} fields when joining many tables.
@item
Fixed a problem with @code{BDB} tables and @code{TEXT} columns.
@item
Fixed bug when using a @code{BLOB} key where a const row wasn't found.
@item
Fixed that @code{mysqlbinlog} writes the timestamp value for each query.
This ensures that one gets same values for date functions like @code{NOW()}
when using @code{mysqlbinlog} to pipe the queries to another server.
@item
Allow @code{--skip-gemini}, @code{--skip-bdb}, and @code{--skip-innodb}
options to be specified when invoking @code{mysqld}, even if these table
handlers are not compiled in to @code{mysqld}.
@item
One can now do @code{GROUP BY ... DESC}.
@item
Fixed a deadlock in the @code{SET} code, when one ran @code{SET @@foo=bar},
where @code{bar} is a column reference, an error was not properly generated.
@end itemize


@node News-3.23.33, News-3.23.32, News-3.23.34, News-3.23.x
@appendixsubsec Changes in release 3.23.33 (09 Feb 2001)

@itemize @bullet
@item
Fixed DNS lookups not to use the same mutex as the hostname
cache.  This will enable known hosts to be quickly resolved even if a
DNS lookup takes a long time.
@item
Added @code{--character-sets-dir} option to @code{myisampack}.
@item
Removed warnings when running @code{REPAIR TABLE ... EXTENDED}.
@item
Fixed a bug that caused a core dump when using @code{GROUP BY} on an alias,
where the alias was the same as an existing column name.
@item
Added @code{SEQUENCE()} as an example UDF function.
@item
Changed @code{mysql_install_db} to use @code{BINARY} for @code{CHAR}
columns in the privilege tables.
@item
Changed @code{TRUNCATE tbl_name} to @code{TRUNCATE TABLE tbl_name}
to use the same syntax as Oracle.  Until 4.0 we will also allow
@code{TRUNCATE tbl_name} to not crash old code.
@item
Fixed ``no found rows'' bug in @code{MyISAM} tables when a @code{BLOB} was
first part of a multi-part key.
@item
Fixed bug where @code{CASE} didn't work with @code{GROUP BY}.
@item
Added @code{--sort-recover} option to @code{myisamchk}.
@item
@code{myisamchk -S} and @code{OPTIMIZE TABLE} now work on Windows.
@item
Fixed bug when using @code{DISTINCT} on results from functions that referred
to a group function, like:
@example
SELECT a, DISTINCT SEC_TO_TIME(SUM(a))
FROM tbl_name GROUP BY a, b;
@end example
@item
Fixed buffer overrun in @code{libmysqlclient} library.
Fixed bug in handling @code{STOP} event after @code{ROTATE} event in
replication.
@item
Fixed another buffer overrun in @code{DROP DATABASE}.
@item
Added @code{Table_locks_immediate} and @code{Table_locks_waited} status
variables.
@item
Fixed bug in replication that broke slave server start with existing
@file{master.info}. This fixes a bug introduced in 3.23.32.
@item
Added @code{SET SQL_SLAVE_SKIP_COUNTER=n} command to recover from
replication glitches without a full database copy.
@item
Added @code{max_binlog_size} variable; the binary log will be rotated
automatically when the size crosses the limit.
@item
Added @code{Last_error}, @code{Last_errno}, and @code{Slave_skip_counter}
variables to @code{SHOW SLAVE STATUS}.
@item
Fixed bug in @code{MASTER_POS_WAIT()} function.
@item
Execute core dump handler on @code{SIGILL}, and @code{SIGBUS} in addition to
@code{SIGSEGV}.
@item
On x86 Linux, print the current query and thread (connection) id, if
available, in the core dump handler.
@item
Fixed several timing bugs in the test suite.
@item
Extended @code{mysqltest} to take care of the timing issues in the test
suite.
@item
@code{ALTER TABLE} can now be used to change the definition for a
@code{MERGE} table.
@item
Fixed creation of @code{MERGE} tables on Windows.
@item
Portability fixes for OpenBSD and OS/2.
@item
Added @code{--temp-pool} option to @code{mysqld}.  Using this option
will cause most temporary files created to use a small set of names,
rather than a unique name for each new file.  This is to work around a
problem in the Linux kernel dealing with creating a bunch of new files
with different names.  With the old behavior, Linux seems to "leak"
memory, as it's being allocated to the directory entry cache instead of
the disk cache.
@end itemize


@node News-3.23.32, News-3.23.31, News-3.23.33, News-3.23.x
@appendixsubsec Changes in release 3.23.32 (22 Jan 2001: Stable)

@itemize @bullet
@item
Changed code to get around compiler bug in Compaq C++ on OSF/1, that broke
@code{BACKUP}, @code{RESTORE}, @code{CHECK}, @code{REPAIR}, and
@code{ANALYZE TABLE}.
@item
Added option @code{FULL} to @code{SHOW COLUMNS}. Now we show the
privilege list for the columns only if this option is given.
@item
Fixed bug in @code{SHOW LOGS} when there weren't any @code{BDB} logs.
@item
Fixed a timing problem in replication that could delay sending an update
to the client until a new update was done.
@item
Don't convert field names when using @code{mysql_list_fields()}.  This is
to keep this code compatible with @code{SHOW FIELDS}.
@item
@code{MERGE} tables didn't work on Windows.
@item
Fixed problem with @code{SET PASSWORD=...} on Windows.
@item
Added missing @file{my_config.h} to RPM distribution.
@item
@code{TRIM("foo" from "foo")} didn't return an empty string.
@item
Added @code{--with-version-suffix} option to @code{configure}.
@item
Fixed core dump when client aborted connection without @code{mysql_close()}.
@item
Fixed a bug in @code{RESTORE TABLE} when trying to restore from a non-existent
directory.
@item
Fixed a bug which caused a core dump on the slave when replicating
@code{SET PASSWORD}.
@item
Added @code{MASTER_POS_WAIT()}.
@end itemize


@node News-3.23.31, News-3.23.30, News-3.23.32, News-3.23.x
@appendixsubsec Changes in release 3.23.31 (17 Jan 2001)

@itemize @bullet
@item
The test suite now tests all reachable @code{BDB} interface code.  During
testing we found and fixed many errors in the interface code.
@item
Using @code{HAVING} on an empty table could produce one result row when
it shouldn't.
@item
Fixed the MySQL RPM not to depend on Perl5 anymore.
@item
Fixed some problems with @code{HEAP} tables on Windows.
@item
@code{SHOW TABLE STATUS} didn't show correct average row length for tables
larger than 4G.
@item
@code{CHECK TABLE ... EXTENDED} didn't check row links for fixed size tables.
@item
Added option @code{MEDIUM} to @code{CHECK TABLE}.
@item
Fixed problem when using @code{DECIMAL()} keys on negative numbers.
@item
@code{HOUR()} (and some other @code{TIME} functions) on a @code{CHAR} column
always returned @code{NULL}.
@item
Fixed security bug in something (please upgrade if you are using an earlier
MySQL 3.23 version).
@item
Fixed buffer overflow bug when writing a certain error message.
@item
Added usage of @code{setrlimit()} on Linux to get
@code{-O --open-files-limit=#} to work on Linux.
@item
Added @code{bdb_version} variable to @code{mysqld}.
@item
Fixed bug when using expression of type:
@example
SELECT ... FROM t1 LEFT JOIN t2 ON (t1.a=t2.a) WHERE t1.a=t2.a
@end example
In this case the test in the @code{WHERE} clause was wrongly optimised away.
@item
Fixed bug in @code{MyISAM} when deleting keys with possible @code{NULL}
values, but the first key-column was not a prefix-compressed text column.
@item
Fixed @code{mysql.server} to read the @code{[mysql.server]} option file group
rather than the @code{[mysql_server]} group.
@item
Fixed @code{safe_mysqld} and @code{mysql.server} to also read the
@code{server} option section.
@item
Added @code{Threads_created} status variable to @code{mysqld}.
@end itemize


@node News-3.23.30, News-3.23.29, News-3.23.31, News-3.23.x
@appendixsubsec Changes in release 3.23.30 (04 Jan 2001)

@itemize @bullet
@item
Added @code{SHOW OPEN TABLES} command.
@item
Fixed that @code{myisamdump} works against old @code{mysqld} servers.
@item
Fixed @code{myisamchk -k#} so that it works again.
@item
Fixed a problem with replication when the binary log file went over 2G
on 32-bit systems.
@item
@code{LOCK TABLES} will now automatically start a new transaction.
@item
Changed @code{BDB} tables to not use internal subtransactions and reuse
open files to get more speed.
@item
Added @code{--mysqld=#} option to @code{safe_mysqld}.
@item
Allow hex constants in the @code{--fields-*-by} and
@code{--lines-terminated-by} options to @code{mysqldump} and
@code{mysqlimport}. By Paul DuBois.
@item
Added @code{--safe-show-database} option to @code{mysqld}.
@item
Added @code{have_bdb}, @code{have_gemini}, @code{have_innobase},
@code{have_raid} and @code{have_openssl} to @code{SHOW VARIABLES} to make it
easy to test for supported extensions.
@item
Added @code{--open-files-limit} option to @code{mysqld}.
@item
Changed @code{--open-files} option to @code{--open-files-limit} in
@code{safe_mysqld}.
@item
Fixed a bug where some rows were not found with @code{HEAP} tables
that had many keys.
@item
Fixed that @code{--bdb-no-sync} works.
@item
Changed @code{--bdb-recover} to @code{--bdb-no-recover} as recover should
be on by default.
@item
Changed the default number of @code{BDB} locks to 10000.
@item
Fixed a bug from 3.23.29 when allocating the shared structure needed
for @code{BDB} tables.
@item
Changed @code{mysqld_multi.sh} to use configure variables. Patch by
Christopher McCrory.
@item
Added fixing of include files for Solaris 2.8.
@item
Fixed bug with @code{--skip-networking} on Debian Linux.
@item
Fixed problem that some temporary files where reported as having the
name @code{UNOPENED} in error messages.
@item
Fixed bug when running two simultaneous @code{SHOW LOGS} queries.
@end itemize


@node News-3.23.29, News-3.23.28, News-3.23.30, News-3.23.x
@appendixsubsec Changes in release 3.23.29 (16 Dec 2000)

@itemize @bullet
@item
Configure updates for Tru64, large file support, and better TCP wrapper
support. By Albert Chin-A-Young.
@item
Fixed bug in @code{<=>} operator.
@item
Fixed bug in @code{REPLACE} with @code{BDB} tables.
@item
@code{LPAD()} and @code{RPAD()} will shorten the result string if it's longer
than the length argument.
@item
Added @code{SHOW LOGS} command.
@item
Remove unused @code{BDB} logs on shutdown.
@item
When creating a table, put @code{PRIMARY} keys first, followed by
@code{UNIQUE} keys.
@item
Fixed a bug in @code{UPDATE} involving multi-part keys where one
specified all key parts both in the update and the @code{WHERE} part. In
this case MySQL could try to update a record that didn't match
the whole @code{WHERE} part.
@item
Changed drop table to first drop the tables and then the @file{.frm} file.
@item
Fixed a bug in the hostname cache which caused @code{mysqld} to report the
hostname as @code{''} in some error messages.
@item
Fixed a bug with @code{HEAP} type tables; the variable
@code{max_heap_table_size} wasn't used. Now either @code{MAX_ROWS} or
@code{max_heap_table_size} can be used to limit the size of a @code{HEAP}
type table.
@item
Changed the default server-id to 1 for masters and 2 for slaves
to make it easier to use the binary log.
@item
Renamed @code{bdb_lock_max} variable to @code{bdb_max_lock}.
@item
Added support for @code{AUTO_INCREMENT} on sub-fields for @code{BDB} tables.
@item
Added @code{ANALYZE} of @code{BDB} tables.
@item
In @code{BDB} tables, we now store the number of rows; this helps to optimise
queries when we need an approximation of the number of rows.
@item
If we get an error in a multi-row statement, we now only rollback the
last statement, not the entire transaction.
@item
If you do a @code{ROLLBACK} when you have updated a non-transactional table
you will get an error as a warning.
@item
Added @code{--bdb-shared-data} option to @code{mysqld}.
@item
Added @code{Slave_open_temp_tables} status variable to @code{mysqld}
@item
Added @code{binlog_cache_size} and @code{max_binlog_cache_size} variables to
@code{mysqld}.
@item
@code{DROP TABLE}, @code{RENAME TABLE}, @code{CREATE INDEX} and
@code{DROP INDEX} are now transaction endpoints.
@item
If you do a @code{DROP DATABASE} on a symbolically linked database, both
the link and the original database is deleted.
@item
Fixed @code{DROP DATABASE} to work on OS/2.
@item
Fixed bug when doing a @code{SELECT DISTINCT ... table1 LEFT JOIN
table2 ...} when @code{table2} was empty.
@item
Added @code{--abort-slave-event-count} and
@code{--disconnect-slave-event-count} options to @code{mysqld} for
debugging and testing of replication.
@item
Fixed replication of temporary tables. Handles everything except
slave server restart.
@item
@code{SHOW KEYS} now shows whether key is @code{FULLTEXT}.
@item
New script @code{mysqld_multi}. @xref{mysqld_multi, , @code{mysqld_multi}}.
@item
Added new script, @code{mysql-multi.server.sh}.  Thanks to
Tim Bunce @email{Tim.Bunce@@ig.co.uk} for modifying @code{mysql.server} to
easily handle hosts running many @code{mysqld} processes.
@item
@code{safe_mysqld}, @code{mysql.server}, and @code{mysql_install_db} have
been modified to use @code{mysql_print_defaults} instead of various hacks
to read the @file{my.cnf} files.  In addition, the handling of various
paths has been made more consistent with how @code{mysqld} handles them
by default.
@item
Automatically remove Berkeley DB transaction logs that no longer are in
use.
@item
Fixed bug with several @code{FULLTEXT} indexes in one table.
@item
Added a warning if number of rows changes on @code{REPAIR}/@code{OPTIMIZE}.
@item
Applied patches for OS/2 by @code{Yuri Dario}.
@item
@code{FLUSH TABLES tbl_name} didn't always flush the index tree
to disk properly.
@item
@code{--bootstrap} is now run in a separate thread. This fixes a problem
that caused @code{mysql_install_db} to core dump on some Linux machines.
@item
Changed @code{mi_create()} to use less stack space.
@item
Fixed bug with optimiser trying to over-optimise @code{MATCH()} when used
with @code{UNIQUE} key.
@item
Changed @code{crash-me} and the MySQL benchmarks to also work
with FrontBase.
@item
Allow @code{RESTRICT} and @code{CASCADE} after @code{DROP TABLE} to make
porting easier.
@item
Reset status variable which could cause problem if one used @code{--slow-log}.
@item
Added @code{connect_timeout} variable to @code{mysql} and @code{mysqladmin}.
@item
Added @code{connect-timeout} as an alias for @code{timeout} for option files
read by @code{mysql_options()}.
@end itemize


@node News-3.23.28, News-3.23.27, News-3.23.29, News-3.23.x
@appendixsubsec Changes in release 3.23.28 (22 Nov 2000: Gamma)

@itemize @bullet
@item
Added new options @code{--pager[=...]}, @code{--no-pager},
@code{--tee=...} and @code{--no-tee} to the @code{mysql} client.  The
new corresponding interactive commands are @code{pager}, @code{nopager},
@code{tee} and @code{notee}. @xref{mysql, , @code{mysql}}, @code{mysql --help}
and the interactive help for more information.
@item
Fixed crash when automatic repair of @code{MyISAM} table failed.
@item
Fixed a major performance bug in the table locking code when one
constantly had a lot of @code{SELECT}, @code{UPDATE} and @code{INSERT}
statements running. The symptom was that the @code{UPDATE} and
@code{INSERT} queries were locked for a long time while new @code{SELECT}
statements were executed before the updates.
@item
When reading @code{options_files} with @code{mysql_options()} the
@code{return-found-rows} option was ignored.
@item
One can now specify @code{interactive-timeout} in the option file that
is read by @code{mysql_options()}. This makes it possible to force
programs that run for a long time (like @code{mysqlhotcopy}) to use the
@code{interactive_timeout} time instead of the @code{wait_timeout} time.
@item
Added to the slow query log the time and the user name for each logged
query. If you are using @code{--log-long-format} then also queries that
do not use an index are logged, even if the query takes less than
@code{long_query_time} seconds.
@item
Fixed a problem in @code{LEFT JOIN} which caused all columns in a reference
table to be @code{NULL}.
@item
Fixed a problem when using @code{NATURAL JOIN} without keys.
@item
Fixed a bug when using a multi-part keys where the first part was of type
@code{TEXT} or @code{BLOB}.
@item
@code{DROP} of temporary tables wasn't stored in the update/binary log.
@item
Fixed a bug where @code{SELECT DISTINCT * ... LIMIT #} only returned one row.
@item
Fixed a bug in the assembler code in @code{strstr()} for SPARC and cleaned up
the @file{global.h} header file to avoid a problem with bad aliasing with
the compiler submitted with RedHat 7.0. (Reported by Trond Eivind Glomsrød)
@item
The @code{--skip-networking} option now works properly on NT.
@item
Fixed a long outstanding bug in the @code{ISAM} tables when a row with a length
of more than 65K was shortened by a single byte.
@item
Fixed a bug in @code{MyISAM} when running multiple updating processes on
the same table.
@item
Allow one to use @code{FLUSH TABLE tbl_name}.
@item
Added @code{--replicate-ignore-table}, @code{--replicate-do-table},
@code{--replicate-wild-ignore-table}, and @code{--replicate-wild-do-table}
options to @code{mysqld}.
@item
Changed all log files to use our own @code{IO_CACHE} mechanism instead of
@code{FILE} to avoid OS problems when there are many files open.
@item
Added @code{--open-files} and @code{--timezone} options to @code{safe_mysqld}.
@item
Fixed a fatal bug in @code{CREATE TEMPORARY TABLE ... SELECT ...}.
@item
Fixed a problem with @code{CREATE TABLE ... SELECT NULL}.
@item
Added variables @code{large_file_support},@code{net_read_timeout},
@code{net_write_timeout} and @code{query_buffer_size} to @code{SHOW VARIABLES}.
@item
Added status variables @code{created_tmp_files} and @code{sort_merge_passes}
to @code{SHOW STATUS}.
@item
Fixed a bug where we didn't allow an index name after the
@code{FOREIGN KEY} definition.
@item
Added @code{TRUNCATE table_name} as a synonym for
@code{DELETE FROM table_name}.
@item
Fixed a bug in a @code{BDB} key compare function when comparing part keys.
@item
Added @code{bdb_lock_max} variable to @code{mysqld}.
@item
Added more tests to the benchmark suite.
@item
Fixed an overflow bug in the client code when using overly long database names.
@item
@code{mysql_connect()} now aborts on Linux if the server doesn't answer in
@code{timeout} seconds.
@item
@code{SLAVE START} did not work if you started with
@code{--skip-slave-start} and had not explicitly run @code{CHANGE MASTER TO}.
@item
Fixed the output of @code{SHOW MASTER STATUS} to be consistent with
@code{SHOW SLAVE STATUS}. (It now has no directory in the log name.)
@item
Added @code{PURGE MASTER LOGS TO}.
@item
Added @code{SHOW MASTER LOGS}.
@item
Added @code{--safemalloc-mem-limit} option to @code{mysqld} to simulate memory
shortage when compiled with the @code{--with-debug=full} option.
@item
Fixed several core dumps in out-of-memory conditions.
@item
@code{SHOW SLAVE STATUS} was using an uninitialised mutex if the slave had
not been started yet.
@item
Fixed bug in @code{ELT()} and @code{MAKE_SET()} when the query used
a temporary table.
@item
@code{CHANGE MASTER TO} without specifying @code{MASTER_LOG_POS} would
set it to 0 instead of 4 and hit the magic number in the master binlog.
@item
@code{ALTER TABLE ... ORDER BY ...} syntax added.  This will create the
new table with the rows in a specific order.
@end itemize


@node News-3.23.27, News-3.23.26, News-3.23.28, News-3.23.x
@appendixsubsec Changes in release 3.23.27 (24 Oct 2000)

@itemize @bullet
@item
Fixed a bug where the automatic repair of @code{MyISAM} tables sometimes failed
when the datafile was corrupt.
@item
Fixed a bug in @code{SHOW CREATE} when using @code{AUTO_INCREMENT} columns.
@item
Changed @code{BDB} tables to use new compare function in Berkeley DB 3.2.3.
@item
You can now use Unix sockets with MIT-pthreads.
@item
Added the @code{latin5} (turkish) character set.
@item
Small portability fixes.
@end itemize


@node News-3.23.26, News-3.23.25, News-3.23.27, News-3.23.x
@appendixsubsec Changes in release 3.23.26

@itemize @bullet
@item
Renamed @code{FLUSH MASTER} and @code{FLUSH SLAVE} to @code{RESET MASTER}
and @code{RESET SLAVE}.
@item
Fixed @code{<>} to work properly with @code{NULL}.
@item
Fixed a problem with @code{SUBSTRING_INDEX()} and @code{REPLACE()}.
(Patch by Alexander Igonitchev)
@item
Fix @code{CREATE TEMPORARY TABLE IF NOT EXISTS} not to produce an error
if the table exists.
@item
If you don't create a @code{PRIMARY KEY} in a @code{BDB} table, a hidden
@code{PRIMARY KEY} will be created.
@item
Added read-only-key optimisation to @code{BDB} tables.
@item
@code{LEFT JOIN} in some cases preferred a full table scan when there was
no @code{WHERE} clause.
@item
When using @code{--log-slow-queries}, don't count the time waiting for a lock.
@item
Fixed bug in lock code on Windows which could cause the key cache
to report that the key file was crashed even if it was okay.
@item
Automatic repair of @code{MyISAM} tables if you start @code{mysqld} with
@code{--myisam-recover}.
@item
Removed the @code{TYPE=} keyword from @code{CHECK} and
@code{REPAIR}. Allow @code{CHECK} options to be combined. (You can still
use @code{TYPE=}, but this usage is deprecated.)
@item
Fixed mutex bug in the binary replication log -- long update queries could
be read only in part by the slave if it did it at the wrong time, which
was not fatal, but resulted in a performance-degrading reconnect and
a scary message in the error log.
@item
Changed the format of the binary log -- added magic number, server
version, binlog version. Added server id and query error code for each
query event.
@item
Replication thread from the slave now will kill all the stale threads
from the same server.
@item
Long replication user names were not being handled properly.
@item
Added @code{--replicate-rewrite-db} option to @code{mysqld}.
@item
Added @code{--skip-slave-start} option to @code{mysqld}.
@item
Updates that generated an error code (such as @code{INSERT INTO foo(some_key)
values (1),(1)}) erroneously terminated the slave thread.
@item
Added optimisation of queries where @code{DISTINCT} is only used on columns
from some of the tables.
@item
Allow floating-point numbers where there is no sign after the exponent
(like @code{1e1}).
@item
@code{SHOW GRANTS} didn't always show all column grants.
@item
Added @code{--default-extra-file=#} option to all MySQL clients.
@item
Columns referenced in @code{INSERT} statements now are initialised properly.
@item
@code{UPDATE} didn't always work when used with a range on a timestamp that
was part of the key that was used to find rows.
@item
Fixed a bug in @code{FULLTEXT} index when inserting a @code{NULL} column.
@item
Changed to use @code{mkstemp()} instead of @code{tempnam()}. Based
on a patch from John Jones.
@end itemize


@node News-3.23.25, News-3.23.24, News-3.23.26, News-3.23.x
@appendixsubsec Changes in release 3.23.25

@itemize @bullet
@item
Fixed that @code{databasename} works as second argument to @code{mysqlhotcopy}.
@item
The values for the @code{UMASK} and @code{UMASK_DIR} environment variables
now can be specified in octal by beginning the value with a zero.
@item
Added @code{RIGHT JOIN}. This makes @code{RIGHT} a reserved word.
@item
Added @code{@@@@IDENTITY} as a synonym for @code{LAST_INSERT_ID()}.
(This is for Visual Basic compatibility.)
@item
Fixed a bug in @code{myisamchk} and @code{REPAIR} when using @code{FULLTEXT}
index.
@item
@code{LOAD DATA INFILE} now works with FIFOs.
(Patch by Toni L. Harbaugh-Blackford.)
@item
@code{FLUSH LOGS} broke replication if you specified a log name with an
explicit extension as the value of the @code{log-bin} option.
@item
Fixed a bug in @code{MyISAM} with packed multi-part keys.
@item
Fixed crash when using @code{CHECK TABLE} on Windows.
@item
Fixed a bug where @code{FULLTEXT} index always used the @code{koi8_ukr}
character set.
@item
Fixed privilege checking for @code{CHECK TABLE}.
@item
The @code{MyISAM} repair/reindex code didn't use the @code{--tmpdir}
option for its temporary files.
@item
Added @code{BACKUP TABLE} and @code{RESTORE TABLE}.
@item
Fixed core dump on @code{CHANGE MASTER TO} when the slave did not have
the master to start with.
@item
Fixed incorrect @code{Time} in the processlist for @code{Connect} of
the slave thread.
@item
The slave now logs when it connects to the master.
@item
Fixed a core dump bug when doing @code{FLUSH MASTER} if you didn't specify
a filename argument to @code{--log-bin}.
@item
Added missing @file{ha_berkeley.x} files to the MySQL Windows distribution.
@item
Fixed some mutex bugs in the log code that could cause thread blocks if new
log files couldn't be created.
@item
Added lock time and number of selected processed rows to slow query log.
@item
Added @code{--memlock} option to @code{mysqld} to lock @code{mysqld}
in memory on systems with the @code{mlockall()} call (like in Solaris).
@item
@code{HEAP} tables didn't use keys properly. (Bug from 3.23.23.)
@item
Added better support for @code{MERGE} tables (keys, mapping, creation,
documentation...). @xref{MERGE, , @code{MERGE}}.
@item
Fixed bug in @code{mysqldump} from 3.23 which caused some @code{CHAR} columns
not to be quoted.
@item
Merged @code{analyze}, @code{check}, @code{optimize} and repair code.
@item
@code{OPTIMIZE TABLE} is now mapped to @code{REPAIR} with statistics and
sorting of the index tree.  This means that for the moment it only
works on @code{MyISAM} tables.
@item
Added a pre-alloced block to root_malloc to get fewer mallocs.
@item
Added a lot of new statistics variables.
@item
Fixed @code{ORDER BY} bug with @code{BDB} tables.
@item
Removed warning that @code{mysqld} couldn't remove the @file{.pid} file
under Windows.
@item
Changed @code{--log-isam} to log @code{MyISAM} tables instead of isam
tables.
@item
Fixed @code{CHECK TABLE} to work on Windows.
@item
Added file mutexes to make @code{pwrite()} safe on Windows.
@end itemize


@node News-3.23.24, News-3.23.23, News-3.23.25, News-3.23.x
@appendixsubsec Changes in release 3.23.24 (08 Sep 2000)

@itemize @bullet
@item
Added @code{created_tmp_disk_tables} variable to @code{mysqld}.
@item
To make it possible to reliably dump and restore tables with
@code{TIMESTAMP(X)} columns, MySQL now reports columns with @code{X}
other than 14 or 8 to be strings.
@item
Changed sort order for @code{latin1} as it was before MySQL Version 3.23.23.
Any table that was created or modified with 3.23.22 must be repaired if it has
@code{CHAR} columns that may contain characters with ASCII values greater than
128!
@item
Fixed small memory leak introduced from 3.23.22 when creating a
temporary table.
@item
Fixed problem with @code{BDB} tables and reading on a unique (not primary) key.
@item
Restored the @code{win1251} character set (it's now only marked deprecated).
@end itemize


@node News-3.23.23, News-3.23.22, News-3.23.24, News-3.23.x
@appendixsubsec Changes in release 3.23.23

@itemize @bullet
@item
Changed sort order for 'German'; all tables created with 'German' sortorder
must be repaired with @code{REPAIR TABLE} or @code{myisamchk} before use!
@item
Added @code{--core-file} option to @code{mysqld} to get a core file on
Linux if @code{mysqld} dies on the @code{SIGSEGV} signal.
@item
MySQL client @code{mysql} now starts with option
@code{--no-named-commands} (@code{-g}) by default. This option can be
disabled with @code{--enable-named-commands} (@code{-G}). This may cause
incompatibility problems in some cases, for example, in SQL scripts that
use named commands without a semicolon, etc. !  Long format commands
still work from the first line.
@item
Fixed a problem when using many pending @code{DROP TABLE} statements at
the same time.
@item
Optimiser didn't use keys properly when using @code{LEFT JOIN} on an
empty table.
@item
Added shorter help text when invoking @code{mysqld} with incorrect options.
@item
Fixed non-fatal @code{free()} bug in @code{mysqlimport}.
@item
Fixed bug in @code{MyISAM} index handling of
@code{DECIMAL}/@code{NUMERIC} keys.
@item
Fixed a bug in concurrent insert in @code{MyISAM} tables.  In some contexts,
usage of @code{MIN(key_part)} or @code{MAX(key_part)} returned an empty set.
@item
Updated @code{mysqlhotcopy} to use the new @code{FLUSH TABLES table_list}
syntax. Only tables which are being backed up are flushed now.
@item
Changed behavior of @code{--enable-thread-safe-client} so
that both non-threaded (@code{-lmysqlclient}) and threaded
(@code{-lmysqlclient_r}) libraries are built.  Users who linked
against a threaded @code{-lmysqlclient} will need to link against
@code{-lmysqlclient_r} now.
@item
Added atomic @code{RENAME TABLE} command.
@item
Don't count @code{NULL} values in @code{COUNT(DISTINCT ...)}.
@item
Changed @code{ALTER TABLE}, @code{LOAD DATA INFILE} on empty tables and
@code{INSERT ... SELECT ...} on empty tables to create non-unique indexes
in a separate batch with sorting. This will make the above calls much
faster when you have many indexes.
@item
@code{ALTER TABLE} now logs the first used insert_id correctly.
@item
Fixed crash when adding a default value to a @code{BLOB} column.
@item
Fixed a bug with @code{DATE_ADD/DATE_SUB} where it returned a datetime instead
of a date.
@item
Fixed a problem with the thread cache which made some threads show up as
@code{***DEAD***} in @code{SHOW PROCESSLIST}.
@item
Fixed a lock in our thr_rwlock code, which could make selects that run
at the same time as concurrent inserts crash. This only affects systems
that don't have the @code{pthread_rwlock_rdlock} code.
@item
When deleting rows with a non-unique key in a @code{HEAP} table, all rows
weren't always deleted.
@item
Fixed bug in range optimiser for @code{HEAP} tables for searches on a part
index.
@item
Fixed @code{SELECT} on part keys to work with @code{BDB} tables.
@item
Fixed @code{INSERT INTO bdb_table ... SELECT} to work with @code{BDB} tables.
@item
@code{CHECK TABLE} now updates key statistics for the table.
@item
@code{ANALYZE TABLE} will now only update tables that have been changed
since the last @code{ANALYZE}. Note that this is a new feature and tables
will not be marked to be analysed until they are updated in any way with
3.23.23 or newer.  For older tables, you have to do @code{CHECK TABLE}
to update the key distribution.
@item
Fixed some minor privilege problems with @code{CHECK}, @code{ANALYZE},
@code{REPAIR} and @code{SHOW CREATE} commands.
@item
Added @code{CHANGE MASTER TO} statement.
@item
Added @code{FAST}, @code{QUICK} @code{EXTENDED} check types to
@code{CHECK TABLES}.
@item
Changed @code{myisamchk} so that @code{--fast} and
@code{--check-only-changed} are also honored with @code{--sort-index} and
@code{--analyze}.
@item
Fixed fatal bug in @code{LOAD TABLE FROM MASTER} that did not lock the
table during index re-build.
@item
@code{LOAD DATA INFILE} broke replication if the database was excluded
from replication.
@item
More variables in @code{SHOW SLAVE STATUS} and @code{SHOW MASTER STATUS}.
@item
@code{SLAVE STOP} now will not return until the slave thread actually exits.
@item
Full-text search via the @code{MATCH()} function and @code{FULLTEXT} index type
(for @code{MyISAM} files).  This makes @code{FULLTEXT} a reserved word.
@end itemize


@node News-3.23.22, News-3.23.21, News-3.23.23, News-3.23.x
@appendixsubsec Changes in release 3.23.22 (31 Jul 2000)

@itemize @bullet
@item
Fixed that @code{lex_hash.h} is created properly for each MySQL
distribution.
@item
Fixed that @code{MASTER} and @code{COLLECTION} are not reserved words.
@item
The log generated by @code{--slow-query-log} didn't contain the whole queries.
@item
Fixed that open transactions in @code{BDB} tables are rolled back if the
connection is closed unexpectedly.
@item
Added workaround for a bug in @code{gcc} 2.96 (intel) and @code{gcc} 2.9
(IA64) in @code{gen_lex_hash.c}.
@item
Fixed memory leak in the client library when using @code{host=} in the
@file{my.cnf} file.
@item
Optimised functions that manipulate the hours/minutes/seconds.
@item
Fixed bug when comparing the result of @code{DATE_ADD()}/@code{DATE_SUB()}
against a number.
@item
Changed the meaning of @code{-F, --fast} for @code{myisamchk}. Added
@code{-C, --check-only-changed} option to @code{myisamchk}.
@item
Added @code{ANALYZE tbl_name} to update key statistics for tables.
@item
Changed binary items @code{0x...} to be regarded as integers by default.
@item
Fix for SCO and @code{SHOW PROCESSLIST}.
@item
Added @code{auto-rehash} on reconnect for the @code{mysql} client.
@item
Fixed a newly introduced bug in @code{MyISAM}, where the index file couldn't
get bigger than 64M.
@item
Added @code{SHOW MASTER STATUS} and @code{SHOW SLAVE STATUS}.
@end itemize


@node News-3.23.21, News-3.23.20, News-3.23.22, News-3.23.x
@appendixsubsec Changes in release 3.23.21

@itemize @bullet
@item
Added @code{mysql_character_set_name()} function to the
MySQL C API.
@item
Made the update log ASCII 0 safe.
@item
Added the @code{mysql_config} script.
@item
Fixed problem when using @code{<} or @code{>} with a char column that was only
partly indexed.
@item
One would get a core dump if the log file was not readable by the MySQL user.
@item
Changed @code{mysqladmin} to use @code{CREATE DATABASE} and @code{DROP
DATABASE} statements instead of the old deprecated API calls.
@item
Fixed @code{chown} warning in @code{safe_mysqld}.
@item
Fixed a bug in @code{ORDER BY} that was introduced in 3.23.19.
@item
Only optimise the @code{DELETE FROM tbl_name} to do a drop+create of
the table if we are in @code{AUTOCOMMIT} mode (needed for @code{BDB} tables).
@item
Added extra checks to avoid index corruption when the @code{ISAM}/@code{MyISAM}
index files get full during an @code{INSERT}/@code{UPDATE}.
@item
@code{myisamchk} didn't correctly update row checksum when used with
@code{-ro} (this only gave a warning in subsequent runs).
@item
Fixed bug in @code{REPAIR TABLE} so that it works with tables without indexes.
@item
Fixed buffer overrun in @code{DROP DATABASE}.
@item
@code{LOAD TABLE FROM MASTER} is sufficiently bug-free to announce it as
a feature.
@item
@code{MATCH} and @code{AGAINST} are now reserved words.
@end itemize


@node News-3.23.20, News-3.23.19, News-3.23.21, News-3.23.x
@appendixsubsec Changes in release 3.23.20

@itemize @bullet
@item
Fixed bug in 3.23.19; @code{DELETE FROM tbl_name} removed the @file{.frm} file.
@item
Added @code{SHOW CREATE TABLE}.
@end itemize


@node News-3.23.19, News-3.23.18, News-3.23.20, News-3.23.x
@appendixsubsec Changes in release 3.23.19

@itemize @bullet
@item
Changed copyright for all files to GPL for the server code and utilities and
to LGPL for the client libraries.
@item
Fixed bug where all rows matching weren't updated on a @code{MyISAM} table
when doing update based on key on a table with many keys and some key changed
values.
@item
The Linux MySQL RPM's and binaries are now statically linked with
a linuxthread version that has faster mutex handling when used with
MySQL.
@item
@code{ORDER BY} can now use @code{REF} keys to find subsets of the rows
that need to be sorted.
@item
Changed name of @code{print_defaults} program to @code{my_print_defaults}
to avoid name confusion.
@item
Fixed @code{NULLIF()} to work according to ANSI SQL99.
@item
Added @code{net_read_timeout} and @code{net_write_timeout} as startup
parameters to @code{mysqld}.
@item
Fixed bug that destroyed index when doing @code{myisamchk --sort-records}
on a table with prefix compressed index.
@item
Added @code{pack_isam} and @code{myisampack} to the standard MySQL
distribution.
@item
Added the syntax @code{BEGIN WORK} (the same as @code{BEGIN}).
@item
Fixed core dump bug when using @code{ORDER BY} on a @code{CONV()} expression.
@item
Added @code{LOAD TABLE FROM MASTER}.
@item
Added @code{FLUSH MASTER} and @code{FLUSH SLAVE}.
@item
Fixed big/little endian problem in the replication.
@end itemize


@node News-3.23.18, News-3.23.17, News-3.23.19, News-3.23.x
@appendixsubsec Changes in release 3.23.18

@itemize @bullet
@item
Fixed a problem from 3.23.17 when choosing character set on the client side.
@item
Added @code{FLUSH TABLES WITH READ LOCK} to make a global lock suitable for
making a copy of MySQL datafiles.
@item
@code{CREATE TABLE ... SELECT ... PROCEDURE} now works.
@item
Internal temporary tables will now use compressed index when using
@code{GROUP BY} on @code{VARCHAR/CHAR} columns.
@item
Fixed a problem when locking the same table with both a @code{READ} and a
@code{WRITE} lock.
@item
Fixed problem with @code{myisamchk} and @code{RAID} tables.
@end itemize


@node News-3.23.17, News-3.23.16, News-3.23.18, News-3.23.x
@appendixsubsec Changes in release 3.23.17

@itemize @bullet
@item
Fixed a bug in @code{FIND_IN_SET()} when the first argument was @code{NULL}.
@item
Added table locks to Berkeley DB.
@item
Fixed a bug with @code{LEFT JOIN} and @code{ORDER BY} where the first
table had only one matching row.
@item
Added 4 sample @file{my.cnf} example files in the @file{support-files}
directory.
@item
Fixed @code{duplicated key} problem when doing big @code{GROUP BY} operations.
(This bug was probably introduced in 3.23.15.)
@item
Changed syntax for @code{INNER JOIN} to match ANSI SQL.
@item
Added @code{NATURAL JOIN} syntax.
@item
A lot of fixes in the @code{BDB} interface.
@item
Added handling of @code{--no-defaults} and @code{--defaults-file} to
@code{safe_mysqld.sh} and @code{mysql_install_db.sh}.
@item
Fixed bug in reading compressed tables with many threads.
@item
Fixed that @code{USE INDEX} works with @code{PRIMARY} keys.
@item
Added @code{BEGIN} statement to start a transaction in @code{AUTOCOMMIT} mode.
@item
Added support for symbolic links for Windows.
@item
Changed protocol to let client know if the server is in @code{AUTOCOMMIT} mode
and if there is a pending transaction.  If there is a pending transaction,
the client library will give an error before reconnecting to the server to
let the client know that the server did a rollback.
The protocol is still backward-compatible with old clients.
@item
@code{KILL} now works on a thread that is locked on a 'write' to a dead client.
@item
Fixed memory leak in the replication slave thread.
@item
Added new @code{log-slave-updates} option to @code{mysqld}, to allow
daisy-chaining the slaves.
@item
Fixed compile error on FreeBSD and other systems where @code{pthread_t}
is not the same as @code{int}.
@item
Fixed master shutdown aborting the slave thread.
@item
Fixed a race condition in @code{INSERT DELAYED} code when doing
@code{ALTER TABLE}.
@item
Added deadlock detection sanity checks to @code{INSERT DELAYED}.
@end itemize


@node News-3.23.16, News-3.23.15, News-3.23.17, News-3.23.x
@appendixsubsec Changes in release 3.23.16

@itemize @bullet
@item
Added @code{SLAVE START} and @code{SLAVE STOP} statements.
@item
Added @code{TYPE=QUICK} option to @code{CHECK} and to @code{REPAIR}.
@item
Fixed bug in @code{REPAIR TABLE} when the table was in use by other threads.
@item
Added a thread cache to make it possible to debug MySQL with
@code{gdb} when one does a lot of reconnects. This will also improve
systems where you can't use persistent connections.
@item
Lots of fixes in the Berkeley DB interface.
@item
@code{UPDATE IGNORE} will not abort if an update results in a
@code{DUPLICATE_KEY} error.
@item
Put @code{CREATE TEMPORARY TABLE} commands in the update log.
@item
Fixed bug in handling of masked IP numbers in the privilege tables.
@item
Fixed bug with @code{delay_key_write} tables and @code{CHECK TABLE}.
@item
Added @code{replicate-do-db} and @code{replicate-ignore-db} options to
@code{mysqld}, to restrict which databases get replicated.
@item
Added @code{SQL_LOG_BIN} option.
@end itemize


@node News-3.23.15, News-3.23.14, News-3.23.16, News-3.23.x
@appendixsubsec Changes in release 3.23.15 (May 2000: Beta)

@itemize @bullet
@item
To start @code{mysqld} as @code{root}, you must now use the @code{--user=root} option.
@item
Added interface to Berkeley DB. (This is not yet functional; play with it at
your own risk!)
@item
Replication between master and slaves.
@item
Fixed bug that other threads could steal a lock when a thread had
a lock on a table and did a @code{FLUSH TABLES} command.
@item
Added the @code{slow_launch_time} variable and the @code{Slow_launch_threads}
status variable to @code{mysqld}.  These can be examined with
@code{mysqladmin variables} and @code{mysqladmin extended-status}.
@item
Added functions @code{INET_NTOA()} and @code{INET_ATON()}.
@item
The default type of @code{IF()} now depends on the second and
third arguments and not only on the second argument.
@item
Fixed case when @code{myisamchk} could go into a loop when trying to
repair a crashed table.
@item
Don't write @code{INSERT DELAYED} to update log if @code{SQL_LOG_UPDATE=0}.
@item
Fixed problem with @code{REPLACE} on @code{HEAP} tables.
@item
Added possible character sets and time zone to @code{SHOW VARIABLES} output.
@item
Fixed bug in locking code that could result in locking problems with
concurrent inserts under high load.
@item
Fixed a problem with @code{DELETE} of many rows on a table with
compressed keys where MySQL scanned the index to find the rows.
@item
Fixed problem with @code{CHECK} on table with deleted keyblocks.
@item
Fixed a bug in reconnect (at the client side) where it didn't free memory
properly in some contexts.
@item
Fixed problems in update log when using @code{LAST_INSERT_ID()} to update
a table with an @code{AUTO_INCREMENT} key.
@item
Added @code{NULLIF()} function.
@item
Fixed bug when using @code{LOAD DATA INFILE} on a table with
@code{BLOB/TEXT} columns.
@item
Optimised @code{MyISAM} to be faster when inserting keys in sorted order.
@item
@code{EXPLAIN SELECT ...} now also prints out whether MySQL needs to
create a temporary table or use file sorting when resolving the @code{SELECT}.
@item
Added optimisation to skip @code{ORDER BY} parts where the part is a
constant expression in the @code{WHERE} part.  Indexes can now be used
even if the @code{ORDER BY} doesn't match the index exactly, as long as
all the unused index parts and all the extra @code{ORDER BY}
columns are constants in the @code{WHERE} clause.  @xref{MySQL indexes}.
@item
@code{UPDATE} and @code{DELETE} on a whole unique key in the @code{WHERE} part
are now faster than before.
@item
Changed @code{RAID_CHUNKSIZE} to be in 1024-byte increments.
@item
Fixed core dump in @code{LOAD_FILE(NULL)}.
@end itemize


@node News-3.23.14, News-3.23.13, News-3.23.15, News-3.23.x
@appendixsubsec Changes in release 3.23.14

@itemize @bullet
@item
Added @code{mysql_real_escape_string()} function to the MySQL C API.
@item
Fixed a bug in @code{CONCAT()} where one of the arguments was a function
that returned a modified argument.
@item
Fixed a critical bug in @code{myisamchk}, where it updated the header in
the index file when one only checked the table.  This confused the
@code{mysqld} daemon if it updated the same table at the same time. Now
the status in the index file is only updated if one uses
@code{--update-state}.  With older @code{myisamchk} versions you should
use @code{--read-only} when only checking tables, if there is the
slightest chance that the @code{mysqld} server is working on the table at the
same time!
@item
Fixed that @code{DROP TABLE} is logged in the update log.
@item
Fixed problem when searching on @code{DECIMAL()} key field
where the column data contained leading zeros.
@item
Fix bug in @code{myisamchk} when the @code{AUTO_INCREMENT} column isn't
the first key.
@item
Allow @code{DATETIME} in ISO8601 format: 2000-03-12T12:00:00
@item
Dynamic character sets.  A @code{mysqld} binary can now handle many different
character sets (you can choose which when starting @code{mysqld}).
@item
Added command @code{REPAIR TABLE}.
@item
Added @code{mysql_thread_safe()} function to the MySQL C API.
@item
Added the @code{UMASK_DIR} environment variable.
@item
Added @code{CONNECTION_ID()} function to return the client connection thread
ID.
@item
When using @code{=} on @code{BLOB} or @code{VARCHAR BINARY} keys, where
only a part of the column was indexed, the whole column of the result
row wasn't compared.
@item
Fix for @code{sjis} character set and @code{ORDER BY}.
@item
When running in ANSI mode, don't allow columns to be used that aren't in
the @code{GROUP BY} part.
@end itemize


@node News-3.23.13, News-3.23.12, News-3.23.14, News-3.23.x
@appendixsubsec Changes in release 3.23.13

@itemize @bullet
@item
Fixed problem when doing locks on the same table more than 2 times in
the same @code{LOCK TABLE} command; this fixed the problem one got when running
the test-ATIS test with @code{--fast} or @code{--check-only-changed}.
@item
Added @code{SQL_BUFFER_RESULT} option to @code{SELECT}.
@item
Removed end space from double/float numbers in results from temporary
tables.
@item
Added @code{CHECK TABLE} command.
@item
Added changes for @code{MyISAM} in 3.23.12 that didn't get into the source
distribution because of CVS problems.
@item
Fixed bug so that @code{mysqladmin shutdown} will wait for the local server
to close down.
@item
Fixed a possible endless loop when calculating timestamp.
@item
Added @code{print_defaults} program to the @file{.rpm} files. Removed
@code{mysqlbug} from the client @file{.rpm} file.
@end itemize


@node News-3.23.12, News-3.23.11, News-3.23.13, News-3.23.x
@appendixsubsec Changes in release 3.23.12

@itemize @bullet
@item
Fixed bug in @code{MyISAM} involving @code{REPLACE ... SELECT ...} which could
give a corrupted table.
@item
Fixed bug in @code{myisamchk} where it incorrectly reset the
@code{AUTO_INCREMENT} value.
@item
LOTS of patches for Linux Alpha. MySQL now appears to be relatively
stable on Alpha.
@item
Changed @code{DISTINCT} on @code{HEAP} temporary tables to use hashed
keys to quickly find duplicated rows. This mostly concerns queries of
type @code{SELECT DISTINCT ... GROUP BY ...}.  This fixes a problem where
not all duplicates were removed in queries of the above type.  In
addition, the new code is MUCH faster.
@item
Added patches to make MySQL compile on Mac OS X.
@item
Added @code{IF NOT EXISTS} clause to @code{CREATE DATABASE}.
@item
Added @code{--all-databases} and @code{--databases} options to @code{mysqldump}
to allow dumping of many databases at the same time.
@item
Fixed bug in compressed @code{DECIMAL()} index in @code{MyISAM} tables.
@item
Fixed bug when storing 0 into a timestamp.
@item
When doing @code{mysqladmin shutdown} on a local connection, @code{mysqladmin}
now waits until the PID file is gone before terminating.
@item
Fixed core dump with some @code{COUNT(DISTINCT ...)} queries.
@item
Fixed that @code{myisamchk} works properly with RAID tables.
@item
Fixed problem with @code{LEFT JOIN} and @code{key_field IS NULL}.
@item
Fixed bug in @code{net_clear()} which could give the error @code{Aborted
connection} in the MySQL clients.
@item
Added options @code{USE INDEX (key_list)} and @code{IGNORE INDEX (key_list)} as
parameters in @code{SELECT}.
@item
@code{DELETE} and @code{RENAME} should now work on @code{RAID} tables.
@end itemize


@node News-3.23.11, News-3.23.10, News-3.23.12, News-3.23.x
@appendixsubsec Changes in release 3.23.11

@itemize @bullet
@item
Allow the @code{ALTER TABLE tbl_name ADD (field_list)} syntax.
@item
Fixed problem with optimiser that could sometimes use incorrect keys.
@item
Fixed that @code{GRANT/REVOKE ALL PRIVILEGES} doesn't affect
@code{GRANT OPTION}.
@item
Removed extra @samp{)} from the output of @code{SHOW GRANTS}.
@item
Fixed problem when storing numbers in timestamps.
@item
Fix problem with timezones that have half hour offsets.
@item
Allow the syntax @code{UNIQUE INDEX} in @code{CREATE} statements.
@item
@code{mysqlhotcopy} - fast online hot-backup utility for local
MySQL databases. By Tim Bunce.
@item
New more secure @code{mysqlaccess}. Thanks to Steve Harvey for this.
@item
Added @code{--i-am-a-dummy} and @code{--safe-updates} options to @code{mysql}.
@item
Added @code{select_limit} and @code{max_join_size} variables to @code{mysql}.
@item
Added @code{SQL_MAX_JOIN_SIZE} and @code{SQL_SAFE_UPDATES} options.
@item
Added @code{READ LOCAL} lock that doesn't lock the table for concurrent
inserts. (This is used by @code{mysqldump}.)
@item
Changed that @code{LOCK TABLES ... READ} doesn't anymore allow concurrent
inserts.
@item
Added @code{--skip-delay-key-write} option to @code{mysqld}.
@item
Fixed security problem in the protocol regarding password checking.
@item
@code{_rowid} can now be used as an alias for an integer type unique indexed
column.
@item
Added back blocking of @code{SIGPIPE} when compiling with @code{--thread-safe-clients}
to make things safe for old clients.
@end itemize


@node News-3.23.10, News-3.23.9, News-3.23.11, News-3.23.x
@appendixsubsec Changes in release 3.23.10

@itemize @bullet
@item
Fixed bug in 3.23.9 where memory wasn't properly freed when using
@code{LOCK TABLES}.
@end itemize


@node News-3.23.9, News-3.23.8, News-3.23.10, News-3.23.x
@appendixsubsec Changes in release 3.23.9

@itemize @bullet
@item
Fixed problem that affected queries that did arithmetic on group functions.
@item
Fixed problem with timestamps and @code{INSERT DELAYED}.
@item
Fixed that @code{date_col BETWEEN const_date AND const_date} works.
@item
Fixed problem when only changing a 0 to @code{NULL} in a table with
@code{BLOB/TEXT} columns.
@item
Fixed bug in range optimiser when using many key parts and or on the middle
key parts:  @code{WHERE K1=1 and K3=2 and (K2=2 and K4=4 or K2=3 and K4=5)}
@item
Added @code{source} command to @code{mysql} to allow reading of batch files
inside the @code{mysql} client.  Original patch by Matthew Vanecek.
@item
Fixed critical problem with the @code{WITH GRANT OPTION} option.
@item
Don't give an unnecessary @code{GRANT} error when using tables from many
databases in the same query.
@item
Added VIO wrapper (needed for SSL support; by Andrei Errapart and
Tõnu Samuel).
@item
Fixed optimiser problem on @code{SELECT} when using many overlapping indexes.
MySQL should now be able to choose keys even better when there
is many keys to choose from.
@item
Changed optimiser to prefer a range key instead of a ref key when the
range key can uses more columns than the ref key (which only can use
columns with @code{=}).  For example, the following type of queries should now
be faster: @code{SELECT * from key_part_1=const and key_part_2 > const2}
@item
Fixed bug that a change of all @code{VARCHAR} columns to @code{CHAR} columns
didn't change row type from dynamic to fixed.
@item
Disabled floating-point exceptions for FreeBSD to fix core dump when
doing @code{SELECT FLOOR(POW(2,63))}.
@item
Renamed @code{mysqld} startup option from @code{--delay-key-write} to
@code{--delay-key-write-for-all-tables}.
@item
Added @code{read-next-on-key} to @code{HEAP} tables.  This should fix all
problems with @code{HEAP} tables when using non-@code{UNIQUE} keys.
@item
Added option to print default arguments to all clients.
@item
Added @code{--log-slow-queries} option to @code{mysqld} to log all queries
that take a long time to a separate log file with a time indicating how
long the query took.
@item
Fixed core dump when doing @code{WHERE key_col=RAND(...)}.
@item
Fixed optimisation bug in @code{SELECT ... LEFT JOIN ... key_col IS NULL},
when @code{key_col} could contain @code{NULL} values.
@item
Fixed problem with 8-bit characters as separators in @code{LOAD DATA INFILE}.
@end itemize


@node News-3.23.8, News-3.23.7, News-3.23.9, News-3.23.x
@appendixsubsec Changes in release 3.23.8

@itemize @bullet
@item
Fixed problem when handling indexfiles larger than 8G.
@item
Added latest patches to MIT-pthreads for NetBSD.
@item
Fixed problem with timezones that are < GMT - 11.
@item
Fixed a bug when deleting packed keys in @code{NISAM}.
@item
Fixed problem with @code{ISAM} when doing some @code{ORDER BY ... DESC} queries.
@item
Fixed bug when doing a join on a text key which didn't cover the whole key.
@item
Option @code{--delay-key-write} didn't enable delayed key writing.
@item
Fixed update of @code{TEXT} column which involved only case changes.
@item
Fixed that @code{INSERT DELAYED} doesn't update timestamps that are given.
@item
Added function @code{YEARWEEK()} and options @code{x}, @code{X}, @code{v} and
@code{V} to @code{DATE_FORMAT()}.
@item
Fixed problem with @code{MAX(indexed_column)} and @code{HEAP} tables.
@item
Fixed problem with @code{BLOB NULL} keys and @code{LIKE "prefix%"}.
@item
Fixed problem with @code{MyISAM} and fixed-length rows < 5 bytes.
@item
Fixed problem that could cause MySQL to touch freed memory when
doing very complicated @code{GROUP BY} queries.
@item
Fixed core dump if you got a crashed table where an @code{ENUM} field value
was too big.
@end itemize


@node News-3.23.7, News-3.23.6, News-3.23.8, News-3.23.x
@appendixsubsec Changes in release 3.23.7

@itemize @bullet
@item
Fixed workaround under Linux to avoid problems with @code{pthread_mutex_timedwait},
which is used with @code{INSERT DELAYED}. @xref{Linux}.
@item
Fixed that one will get a 'disk full' error message if one gets disk full
when doing sorting (instead of waiting until we got more disk space).
@item
Fixed a bug in @code{MyISAM} with keys > 250 characters.
@item
In @code{MyISAM} one can now do an @code{INSERT} at the same time as other
threads are reading from the table.
@item
Added @code{max_write_lock_count} variable to @code{mysqld} to force a
@code{READ} lock after a certain number of @code{WRITE} locks.
@item
Inverted flag @code{delay_key_write} on @code{show variables}.
@item
Renamed @code{concurrency} variable to @code{thread_concurrency}.
@item
The following functions are now multi-byte-safe:
@code{LOCATE(substr,str)}, @code{POSITION(substr IN str)},
@code{LOCATE(substr,str,pos)}, @code{INSTR(str,substr)},
@code{LEFT(str,len)}, @code{RIGHT(str,len)},
@code{SUBSTRING(str,pos,len)}, @code{SUBSTRING(str FROM pos FOR len)},
@code{MID(str,pos,len)}, @code{SUBSTRING(str,pos)}, @code{SUBSTRING(str
FROM pos)}, @code{SUBSTRING_INDEX(str,delim,count)}, @code{RTRIM(str)},
@code{TRIM([[BOTH | TRAILING] [remstr] FROM] str)},
@code{REPLACE(str,from_str,to_str)}, @code{REVERSE(str)},
@code{INSERT(str,pos,len,newstr)}, @code{LCASE(str)}, @code{LOWER(str)},
@code{UCASE(str)} and @code{UPPER(str)}; patch by Wei He.
@item
Fix core dump when releasing a lock from a non-existent table.
@item
Remove locks on tables before starting to remove duplicates.
@item
Added option @code{FULL} to @code{SHOW PROCESSLIST}.
@item
Added option @code{--verbose} to @code{mysqladmin}.
@item
Fixed problem when automatically converting @code{HEAP} to @code{MyISAM}.
@item
Fixed bug in @code{HEAP} tables when doing insert + delete + insert + scan the
table.
@item
Fixed bugs on Alpha with @code{REPLACE()} and @code{LOAD DATA INFILE}.
@item
Added @code{interactive_timeout} variable to @code{mysqld}.
@item
Changed the argument to @code{mysql_data_seek()} from @code{ulong} to
@code{ulonglong}.
@end itemize


@node News-3.23.6, News-3.23.5, News-3.23.7, News-3.23.x
@appendixsubsec Changes in release 3.23.6

@itemize @bullet
@item
Added @code{-O lower_case_table_names=@{0|1@}} option to @code{mysqld} to allow
users to force table names to lowercase.
@item
Added @code{SELECT ... INTO DUMPFILE}.
@item
Added @code{--ansi} option to @code{mysqld} to make some functions
ANSI SQL compatible.
@item
Temporary table names now start with @code{#sql}.
@item
Added quoting of identifiers with @code{`} (@code{"} in @code{--ansi} mode).
@item
Changed to use @code{snprintf()} when printing floats to avoid some buffer
overflows on FreeBSD.
@item
Made @code{FLOOR()} overflow safe on FreeBSD.
@item
Added @code{--quote-names} option to @code{mysqldump}.
@item
Fixed bug that one could make a part of a @code{PRIMARY KEY NOT NULL}.
@item
Fixed @code{encrypt()} to be thread-safe and not reuse buffer.
@item
Added @code{mysql_odbc_escape_string()} function to support big5 characters in
MyODBC.
@item
Rewrote the table handler to use classes.  This introduces a lot of new code,
but will make table handling faster and better.
@item
Added patch by Sasha for user-defined variables.
@item
Changed that @code{FLOAT} and @code{DOUBLE} (without any length modifiers)
no longer are fixed decimal point numbers.
@item
Changed the meaning of @code{FLOAT(X)}: Now this is the same as @code{FLOAT} if
@code{X} <= 24 and a @code{DOUBLE} if 24 < @code{X} <= 53.
@item
@code{DECIMAL(X)} is now an alias for @code{DECIMAL(X,0)} and @code{DECIMAL}
is now an alias for @code{DECIMAL(10,0)}.  The same goes for @code{NUMERIC}.
@item
Added option @code{ROW_FORMAT=@{default | dynamic | static | compressed@}} to
@code{CREATE_TABLE}.
@item
@code{DELETE FROM table_name} didn't work on temporary tables.
@item
Changed function @code{CHAR_LENGTH()} to be multi-byte character safe.
@item
Added function @code{ORD(string)}.
@end itemize


@node News-3.23.5, News-3.23.4, News-3.23.6, News-3.23.x
@appendixsubsec Changes in release 3.23.5

@itemize @bullet
@item
Fixed some Y2K problems in the new date handling in 3.23.
@item
Fixed problem with @code{SELECT DISTINCT ... ORDER BY RAND()}.
@item
Added patches by Sergei A. Golubchik for text searching on the @code{MyISAM}
level.
@item
Fixed cache overflow problem when using full joins without keys.
@item
Fixed some configure issues.
@item
Some small changes to make parsing faster.
@item
Adding a column after the last field with @code{ALTER TABLE} didn't work.
@item
Fixed problem when using an @code{AUTO_INCREMENT} column in two keys
@item
With @code{MyISAM}, you now can have an @code{AUTO_INCREMENT} column as a key
sub part:
@code{CREATE TABLE foo (a INT NOT NULL AUTO_INCREMENT, b CHAR(5), PRIMARY KEY (b,a))}
@item
Fixed bug in @code{MyISAM} with packed char keys that could be @code{NULL}.
@item
@code{AS} on field name with @code{CREATE TABLE table_name SELECT ...} didn't
work.
@item
Allow use of @code{NATIONAL} and @code{NCHAR} when defining character columns.
This is the same as not using @code{BINARY}.
@item
Don't allow @code{NULL} columns in a @code{PRIMARY KEY} (only in @code{UNIQUE}
keys).
@item
Clear @code{LAST_INSERT_ID()} if one uses this in ODBC:
@code{WHERE auto_increment_column IS NULL}.  This seems to fix some problems
with Access.
@item
@code{SET SQL_AUTO_IS_NULL=0|1} now turns on/off the handling of
searching after the last inserted row with @code{WHERE
auto_increment_column IS NULL}.
@item
Added new variable @code{concurrency} to @code{mysqld} for Solaris.
@item
Added @code{--relative} option to @code{mysqladmin} to make
@code{extended-status} more useful to monitor changes.
@item
Fixed bug when using @code{COUNT(DISTINCT ...)} on an empty table.
@item
Added support for the Chinese character set GBK.
@item
Fixed problem with @code{LOAD DATA INFILE} and @code{BLOB} columns.
@item
Added bit operator @code{~} (negation).
@item
Fixed problem with @code{UDF} functions.
@end itemize


@node News-3.23.4, News-3.23.3, News-3.23.5, News-3.23.x
@appendixsubsec Changes in release 3.23.4

@itemize @bullet
@item
Inserting a @code{DATETIME} into a @code{TIME} column no longer will
try to store 'days' in it.
@item
Fixed problem with storage of float/double on little endian machines.
(This affected @code{SUM()}.)
@item
Added connect timeout on TCP/IP connections.
@item
Fixed problem with @code{LIKE "%"} on an index that may have @code{NULL} values.
@item
@code{REVOKE ALL PRIVILEGES} didn't revoke all privileges.
@item
Allow creation of temporary tables with same name as the original table.
@item
When granting a user a @strong{grant} option for a database, he couldn't grant
privileges to other users.
@item
New command: @code{SHOW GRANTS FOR user} (by Sinisa).
@item
New @code{date_add} syntax:  @code{date/datetime + INTERVAL # interval_type}.
By Joshua Chamas.
@item
Fixed privilege check for @code{LOAD DATA REPLACE}.
@item
Automatic fixing of broken include files on Solaris 2.7
@item
Some configure issues to fix problems with big filesystem detection.
@item
@code{REGEXP} is now case-insensitive if you use non-binary strings.
@end itemize


@node News-3.23.3, News-3.23.2, News-3.23.4, News-3.23.x
@appendixsubsec Changes in release 3.23.3

@itemize @bullet
@item
Added patches for MIT-pthreads on NetBSD.
@item
Fixed range bug in @code{MyISAM}.
@item
@code{ASC} is now the default again for @code{ORDER BY}.
@item
Added @code{LIMIT} to @code{UPDATE}.
@item
Added @code{mysql_change_user()} function to the MySQL C API.
@item
Added character set to @code{SHOW VARIABLES}.
@item
Added support of @code{--[whitespace]} comments.
@item
Allow @code{INSERT into tbl_name VALUES ()}, that is, you may now specify
an empty value list to insert a row in which each column is set to its
default value.
@item
Changed @code{SUBSTRING(text FROM pos)} to conform to ANSI SQL. (Before this
construct returned the rightmost @code{pos} characters.)
@item
@code{SUM()} with @code{GROUP BY} returned 0 on some systems.
@item
Changed output for @code{SHOW TABLE STATUS}.
@item
Added @code{DELAY_KEY_WRITE} option to @code{CREATE TABLE}.
@item
Allow @code{AUTO_INCREMENT} on any key part.
@item
Fixed problem with @code{YEAR(NOW())} and @code{YEAR(CURDATE())}.
@item
Added @code{CASE} construct.
@item
New function @code{COALESCE()}.
@end itemize


@node News-3.23.2, News-3.23.1, News-3.23.3, News-3.23.x
@appendixsubsec Changes in release 3.23.2

@itemize @bullet
@item
Fixed range optimiser bug: @code{SELECT * FROM table_name WHERE
key_part1 >= const AND (key_part2 = const OR key_part2 = const)}.  The
bug was that some rows could be duplicated in the result.
@item
Running @code{myisamchk} without @code{-a} updated the index
distribution incorrectly.
@item
@code{SET SQL_LOW_PRIORITY_UPDATES=1} was causing a parse error.
@item
You can now update index columns that are used in the @code{WHERE} clause.
@code{UPDATE tbl_name SET KEY=KEY+1 WHERE KEY > 100}
@item
Date handling should now be a bit faster.
@item
Added handling of fuzzy dates (dates where day or month is 0), such as
@code{'1999-01-00'}.
@item
Fixed optimisation of @code{SELECT ... WHERE key_part1=const1 AND
key_part_2=const2 AND key_part1=const4 AND key_part2=const4}; indextype
should be @code{range} instead of @code{ref}.
@item
Fixed @code{egcs} 1.1.2 optimiser bug (when using @code{BLOB}s) on Linux Alpha.
@item
Fixed problem with @code{LOCK TABLES} combined with @code{DELETE FROM table}.
@item
@code{MyISAM} tables now allow keys on @code{NULL} and @code{BLOB/TEXT} columns.
@item
The following join is now much faster:
@code{SELECT ... FROM t1 LEFT JOIN t2 ON ... WHERE t2.not_null_column IS NULL}.
@item
@code{ORDER BY} and @code{GROUP BY} can be done on functions.
@item
Changed handling of 'const_item' to allow handling of @code{ORDER BY RAND()}.
@item
Indexes are now used for @code{WHERE key_column = function}.
@item
Indexes are now used for @code{WHERE key_column = col_name} even if
the columns are not identically packed.
@item
Indexes are now used for @code{WHERE col_name IS NULL}.
@item
Changed heap tables to be stored in low_byte_first order (to make it easy
to convert to @code{MyISAM} tables)
@item
Automatic change of @code{HEAP} temporary tables to @code{MyISAM} tables
in case of 'table is full' errors.
@item
Added @code{--init-file=file_name} option to @code{mysqld}.
@item
Added @code{COUNT(DISTINCT value, [value, ...])}.
@item
@code{CREATE TEMPORARY TABLE} now creates a temporary table, in its own
namespace, that is automatically deleted if connection is dropped.
@item
New reserved words (required for @code{CASE}):  @code{CASE, THEN, WHEN, ELSE and END}.
@item
New functions @code{EXPORT_SET()} and @code{MD5()}.
@item
Support for the GB2312 Chinese character set.
@end itemize


@node News-3.23.1, News-3.23.0, News-3.23.2, News-3.23.x
@appendixsubsec Changes in release 3.23.1

@itemize @bullet
@item
Fixed some compilation problems.
@end itemize


@node News-3.23.0,  , News-3.23.1, News-3.23.x
@appendixsubsec Changes in release 3.23.0 (Sep 1999: Alpha)

@itemize @bullet
@item
A new table handler library (@code{MyISAM}) with a lot of new features.
@xref{MyISAM, , @code{MyISAM}}.
@item
You can create in-memory @code{HEAP} tables which are extremely fast for
lookups.
@item
Support for big files (63-bit) on OSes that support big files.
@item
New function @code{LOAD_FILE(filename)} to get the contents of a file as a
string value.
@item
New operator @code{<=>} which will act as @code{=} but will return TRUE
if both arguments are @code{NULL}.  This is useful for comparing changes
between tables.
@item
Added the ODBC 3.0 @code{EXTRACT(interval FROM datetime)} function.
@item
Columns defined as @code{FLOAT(X)} are not rounded on storage and may be
in scientific notation (1.0 E+10) when retrieved.
@item
@code{REPLACE} is now faster than before.
@item
Changed @code{LIKE} character comparison to behave as @code{=};
This means that @code{'e' LIKE '@'e'} is now true.  (If the line doesn't
display correctly, the latter 'e' is a French 'e' with a dot above.)
@item
@code{SHOW TABLE STATUS} returns a lot of information about the tables.
@item
Added @code{LIKE} to the @code{SHOW STATUS} command.
@item
Added @code{Privileges} column to @code{SHOW COLUMNS}.
@item
Added @code{Packed} and @code{Comment} columns to @code{SHOW INDEX}.
@item
Added comments to tables (with @code{CREATE TABLE ... COMMENT "xxx"}).
@item
Added @code{UNIQUE}, as in
@code{CREATE TABLE table_name (col int not null UNIQUE)}
@item
New create syntax: @code{CREATE TABLE table_name SELECT ...}
@item
New create syntax: @code{CREATE TABLE IF NOT EXISTS ...}
@item
Allow creation of @code{CHAR(0)} columns.
@item
@code{DATE_FORMAT()} now requires @samp{%} before any format character.
@item
@code{DELAYED} is now a reserved word (sorry about that :( ).
@item
An example procedure is added: @code{analyse}, file: @file{sql_analyse.c}.
This will describe the data in your query.  Try the following:
@example
SELECT ... FROM ...
WHERE ... PROCEDURE ANALYSE([max elements,[max memory]])
@end example

This procedure is extremely useful when you want to check the data in your
table!
@item
@code{BINARY} cast to force a string to be compared in case-sensitive fashion.
@item
Added @code{--skip-show-database} option to @code{mysqld}.
@item
Check whether a row has changed in an @code{UPDATE} now also works
with @code{BLOB}/@code{TEXT} columns.
@item
Added the @code{INNER} join syntax.  @strong{NOTE}:  This made @code{INNER}
a reserved word!
@item
Added support for netmasks to the hostname in the MySQL grant tables.
You can specify a netmask using the @code{IP/NETMASK} syntax.
@item
If you compare a @code{NOT NULL DATE/DATETIME} column with @code{IS
NULL}, this is changed to a compare against @code{0} to satisfy some ODBC
applications.  (By @email{shreeve@@uci.edu}.)
@item
@code{NULL IN (...)} now returns @code{NULL} instead of @code{0}.  This will
ensure that @code{null_column NOT IN (...)} doesn't match
@code{NULL} values.
@item
Fix storage of floating-point values in @code{TIME} columns.
@item
Changed parsing of @code{TIME} strings to be more strict.  Now the
fractional second part is detected (and currently skipped). The
following formats are supported:
@itemize
@item [[DAYS] [H]H:]MM:]SS[.fraction]
@item [[[[[H]H]H]H]MM]SS[.fraction]
@end itemize
@item
Detect (and ignore) fractional second part from @code{DATETIME}.
@item
Added the @code{LOW_PRIORITY} attribute to @code{LOAD DATA INFILE}.
@item
The default index name now uses the same case as the column name on which the
index name is based.
@item
Changed default number of connections to 100.
@item
Use bigger buffers when using @code{LOAD DATA INFILE}.
@item
@code{DECIMAL(x,y)} now works according to ANSI SQL.
@item
Added aggregate UDF functions. Thanks to Andreas F. Bobak
(@email{bobak@@relog.ch}) for this!
@item
@code{LAST_INSERT_ID()} is now updated for @code{INSERT INTO ... SELECT}.
@item
Some small changes to the join table optimiser to make some joins faster.
@item
@code{SELECT DISTINCT} is much faster; it uses the new @code{UNIQUE}
functionality in @code{MyISAM}. One difference compared to MySQL Version 3.22
is that the output of @code{DISTINCT} is not sorted anymore.
@item
All C client API macros are now functions to make shared libraries more
reliable. Because of this, you can no longer call @code{mysql_num_fields()} on
a @code{MYSQL} object, you must use @code{mysql_field_count()} instead.
@item
Added use of @code{LIBEWRAP}; patch by Henning P. Schmiedehausen.
@item
Don't allow @code{AUTO_INCREMENT} for other than numerical columns.
@item
Using @code{AUTO_INCREMENT} will now automatically make the column
@code{NOT NULL}.
@item
Show @code{NULL} as the default value for @code{AUTO_INCREMENT} columns.
@item
Added @code{SQL_BIG_RESULT}; @code{SQL_SMALL_RESULT} is now default.
@item
Added a shared library RPM. This enhancement was contributed by David
Fox (@email{dsfox@@cogsci.ucsd.edu}).
@item
Added  @code{--enable-large-files} and @code{--disable-large-files} switches
to @code{configure}. See @file{configure.in} for some systems where this is
automatically turned off because of broken implementations.
@item
Upgraded @code{readline} to 4.0.
@item
New @code{CREATE TABLE} options: @code{PACK_KEYS} and @code{CHECKSUM}.
@item
Added @code{--default-table-type} option to @code{mysqld}.
@end itemize


@node News-3.22.x, News-3.21.x, News-3.23.x, News
@appendixsec Changes in release 3.22.x (Older; still supported)

@cindex changes, version 3.22

The 3.22 version has faster and safer connect code than version 3.21, as well
as a lot of new nice enhancements.
As there aren't really any major changes, upgrading from 3.21 to 3.22 should
be very easy and painless.  @xref{Upgrading-from-3.21}.

@menu
* News-3.22.35::                Changes in release 3.22.35
* News-3.22.34::                Changes in release 3.22.34
* News-3.22.33::                Changes in release 3.22.33
* News-3.22.32::                Changes in release 3.22.32
* News-3.22.31::                Changes in release 3.22.31
* News-3.22.30::                Changes in release 3.22.30
* News-3.22.29::                Changes in release 3.22.29
* News-3.22.28::                Changes in release 3.22.28
* News-3.22.27::                Changes in release 3.22.27
* News-3.22.26::                Changes in release 3.22.26
* News-3.22.25::                Changes in release 3.22.25
* News-3.22.24::                Changes in release 3.22.24
* News-3.22.23::                Changes in release 3.22.23
* News-3.22.22::                Changes in release 3.22.22
* News-3.22.21::                Changes in release 3.22.21
* News-3.22.20::                Changes in release 3.22.20
* News-3.22.19::                Changes in release 3.22.19 (Mar 1999: Stable)
* News-3.22.18::                Changes in release 3.22.18
* News-3.22.17::                Changes in release 3.22.17
* News-3.22.16::                Changes in release 3.22.16 (Feb 1999: Gamma)
* News-3.22.15::                Changes in release 3.22.15
* News-3.22.14::                Changes in release 3.22.14
* News-3.22.13::                Changes in release 3.22.13
* News-3.22.12::                Changes in release 3.22.12
* News-3.22.11::                Changes in release 3.22.11
* News-3.22.10::                Changes in release 3.22.10
* News-3.22.9::                 Changes in release 3.22.9
* News-3.22.8::                 Changes in release 3.22.8
* News-3.22.7::                 Changes in release 3.22.7 (Sep 1998: Beta)
* News-3.22.6::                 Changes in release 3.22.6
* News-3.22.5::                 Changes in release 3.22.5
* News-3.22.4::                 Changes in release 3.22.4
* News-3.22.3::                 Changes in release 3.22.3
* News-3.22.2::                 Changes in release 3.22.2
* News-3.22.1::                 Changes in release 3.22.1 (Jun 1998: Alpha)
* News-3.22.0::                 Changes in release 3.22.0
@end menu


@node News-3.22.35, News-3.22.34, News-3.22.x, News-3.22.x
@appendixsubsec Changes in release 3.22.35

@itemize @bullet
@item
Fixed problem with @code{STD()}.
@item
Merged changes from the newest @code{ISAM} library from 3.23.
@item
Fixed problem with @code{INSERT DELAYED}.
@item
Fixed a bug core dump when using a @code{LEFT JOIN}/@code{STRAIGHT_JOIN}
on a table with only one row.
@end itemize


@node News-3.22.34, News-3.22.33, News-3.22.35, News-3.22.x
@appendixsubsec Changes in release 3.22.34

@itemize @bullet
@item
Fixed problem with @code{GROUP BY} on @code{TINYBLOB} columns; this
caused bugzilla to not show rows in some queries.
@item
Had to do total recompile of the Windows binary version as VC++ didn't
compile all relevant files for 3.22.33 :(
@end itemize


@node News-3.22.33, News-3.22.32, News-3.22.34, News-3.22.x
@appendixsubsec Changes in release 3.22.33

@itemize @bullet
@item
Fixed problems in Windows when locking tables with @code{LOCK TABLE}.
@item
Quicker kill of @code{SELECT DISTINCT} queries.
@end itemize


@node News-3.22.32, News-3.22.31, News-3.22.33, News-3.22.x
@appendixsubsec Changes in release 3.22.32

@itemize @bullet
@item
Fixed problem when storing numbers in timestamps.
@item
Fix problem with timezones that have half hour offsets.
@item
Added @code{mysqlhotcopy}, a fast online hot-backup utility for local MySQL
databases. By Tim Bunce.
@item
New more secure @code{mysqlaccess}. Thanks to Steve Harvey for this.
@item
Fixed security problem in the protocol regarding password checking.
@item
Fixed problem that affected queries that did arithmetic on @code{GROUP} functions.
@item
Fixed a bug in the @code{ISAM} code when deleting rows on tables with
packed indexes.
@end itemize


@node News-3.22.31, News-3.22.30, News-3.22.32, News-3.22.x
@appendixsubsec Changes in release 3.22.31

@itemize @bullet
@item
A few small fixes for the Windows version.
@end itemize


@node News-3.22.30, News-3.22.29, News-3.22.31, News-3.22.x
@appendixsubsec Changes in release 3.22.30

@itemize @bullet
@item
Fixed optimiser problem on @code{SELECT} when using many overlapping indexes.
@item
Disabled floating-point exceptions for FreeBSD to fix core dump when
doing @code{SELECT FLOOR(POW(2,63))}.
@item
Added print of default arguments options to all clients.
@item
Fixed critical problem with the @code{WITH GRANT OPTION} option.
@item
Fixed non-critical Y2K problem when writing short date to log files.
@end itemize


@node News-3.22.29, News-3.22.28, News-3.22.30, News-3.22.x
@appendixsubsec Changes in release 3.22.29

@itemize @bullet
@item
Upgraded the configure and include files to match the latest 3.23 version.
This should increase portability and make it easier to build shared libraries.
@item
Added latest patches to MIT-pthreads for NetBSD.
@item
Fixed problem with timezones that are < GMT -11.
@item
Fixed a bug when deleting packed keys in NISAM.
@item
Fixed problem that could cause MySQL to touch freed memory when
doing very complicated @code{GROUP BY} queries.
@item
Fixed core dump if you got a crashed table where an @code{ENUM} field value
was too big.
@item
Added @code{mysqlshutdown.exe} and @code{mysqlwatch.exe} to the Windows
distribution.
@item
Fixed problem when doing @code{ORDER BY} on a reference key.
@item
Fixed that @code{INSERT DELAYED} doesn't update timestamps that are given.
@end itemize


@node News-3.22.28, News-3.22.27, News-3.22.29, News-3.22.x
@appendixsubsec Changes in release 3.22.28

@itemize @bullet
@item
Fixed problem with @code{LEFT JOIN} and @code{COUNT()} on a column which
was declared @code{NULL} + and it had a @code{DEFAULT} value.
@item
Fixed core dump problem when using @code{CONCAT()} in a @code{WHERE} clause.
@item
Fixed problem with @code{AVG()} and @code{STD()} with @code{NULL} values.
@end itemize


@node News-3.22.27, News-3.22.26, News-3.22.28, News-3.22.x
@appendixsubsec Changes in release 3.22.27

@itemize @bullet
@item
Fixed prototype in @file{my_ctype.h} when using other character sets.
@item
Some configure issues to fix problems with big filesystem detection.
@item
Fixed problem when sorting on big @code{BLOB} columns.
@item
@code{ROUND()} will now work on Windows.
@end itemize


@node News-3.22.26, News-3.22.25, News-3.22.27, News-3.22.x
@appendixsubsec Changes in release 3.22.26

@itemize @bullet
@item
Fixed core dump with empty @code{BLOB/TEXT} column argument to
@code{REVERSE()}.
@item
Extended @code{/*! */} with version numbers.
@item
Changed @code{SUBSTRING(text FROM pos)} to conform to ANSI SQL. (Before this
construct returned the rightmost 'pos' characters.)
@item
Fixed problem with @code{LOCK TABLES} combined with @code{DELETE FROM table}
@item
Fixed problem that @code{INSERT ... SELECT} didn't use @code{SQL_BIG_TABLES}.
@item
@code{SET SQL_LOW_PRIORITY_UPDATES=#} didn't work.
@item
Password wasn't updated correctly if privileges didn't change on:
@code{GRANT ... IDENTIFIED BY}
@item
Fixed range optimiser bug in
@code{SELECT * FROM table_name WHERE key_part1 >= const AND (key_part2 = const
OR key_part2 = const)}.
@item
Fixed bug in compression key handling in @code{ISAM}.
@end itemize


@node News-3.22.25, News-3.22.24, News-3.22.26, News-3.22.x
@appendixsubsec Changes in release 3.22.25

@itemize @bullet
@item
Fixed some small problems with the installation.
@end itemize


@node News-3.22.24, News-3.22.23, News-3.22.25, News-3.22.x
@appendixsubsec Changes in release 3.22.24

@itemize @bullet
@item
@code{DATA} is not a reserved word anymore.
@item
Fixed optimiser bug with tables with only one row.
@item
Fixed bug when using @code{LOCK TABLES table_name READ; FLUSH TABLES;}
@item
Applied some patches for HP-UX.
@item
@code{isamchk} should now work on Windows.
@item
Changed @file{configure} to not use big file handling on Linux as this
crashes some RedHat 6.0 systems
@end itemize


@node News-3.22.23, News-3.22.22, News-3.22.24, News-3.22.x
@appendixsubsec Changes in release 3.22.23

@itemize @bullet
@item
Upgraded to use Autoconf 2.13, Automake 1.4 and @code{libtool} 1.3.2.
@item
Better support for SCO in @code{configure}.
@item
Added option @code{--defaults-file=###} to option file handling to force use
of only one specific option file.
@item
Extended @code{CREATE} syntax to ignore MySQL Version 3.23 keywords.
@item
Fixed deadlock problem when using @code{INSERT DELAYED} on a table locked with
@code{LOCK TABLES}.
@item
Fixed deadlock problem when using @code{DROP TABLE} on a table that was
locked by another thread.
@item
Add logging of @code{GRANT/REVOKE} commands in the update log.
@item
Fixed @code{isamchk} to detect a new error condition.
@item
Fixed bug in @code{NATURAL LEFT JOIN}.
@end itemize


@node News-3.22.22, News-3.22.21, News-3.22.23, News-3.22.x
@appendixsubsec Changes in release 3.22.22

@itemize @bullet
@item
Fixed problem in the C API when you called @code{mysql_close()} directly after
@code{mysql_init()}.
@item
Better client error message when you can't open socket.
@item
Fixed @code{delayed_insert_thread} counting when you couldn't create a new
delayed_insert thread.
@item
Fixed bug in @code{CONCAT()} with many arguments.
@item
Added patches for DEC 3.2 and SCO.
@item
Fixed path-bug when installing MySQL as a service on NT.
@item
The MySQL-Windows version is now compiled with VC++ 6.0 instead of
with VC++ 5.0.
@item
New installation setup for MySQL-Windows.
@end itemize


@node News-3.22.21, News-3.22.20, News-3.22.22, News-3.22.x
@appendixsubsec Changes in release 3.22.21

@itemize @bullet
@item
Fixed problem with @code{DELETE FROM TABLE} when table was locked by
another thread.
@item
Fixed bug in @code{LEFT JOIN} involving empty tables.
@item
Changed the @code{mysql.db} column from @code{CHAR(32)} to @code{CHAR(60)}.
@item
@code{MODIFY} and @code{DELAYED} are not reserved words anymore.
@item
Fixed a bug when storing days in a @code{TIME} column.
@item
Fixed a problem with @code{Host '...' is not allowed to connect to this MySQL
server} after one had inserted a new MySQL user with a @code{GRANT}
command.
@item
Changed to use @code{TCP_NODELAY} also on Linux (should give faster TCP/IP
connections).
@end itemize


@node News-3.22.20, News-3.22.19, News-3.22.21, News-3.22.x
@appendixsubsec Changes in release 3.22.20

@itemize @bullet
@item
Fixed @code{STD()} for big tables when result should be 0.
@item
The update log didn't have newlines on some operating systems.
@item
@code{INSERT DELAYED} had some garbage at end in the update log.
@end itemize


@node News-3.22.19, News-3.22.18, News-3.22.20, News-3.22.x
@appendixsubsec Changes in release 3.22.19 (Mar 1999: Stable)

@itemize @bullet
@item
Fixed bug in @code{mysql_install_db} (from 3.22.17).
@item
Changed default key cache size to 8M.
@item
Fixed problem with queries that needed temporary tables with @code{BLOB}
columns.
@end itemize


@node News-3.22.18, News-3.22.17, News-3.22.19, News-3.22.x
@appendixsubsec Changes in release 3.22.18

@itemize @bullet
@item
Fixes a fatal problem in 3.22.17 on Linux; after @code{shutdown}
not all threads died properly.
@item
Added option @code{-O flush_time=#} to @code{mysqld}.  This is mostly
useful on Windows and tells how often MySQL should close all
unused tables and flush all updated tables to disk.
@item
Fixed problem that a @code{VARCHAR} column compared with @code{CHAR} column
didn't use keys efficiently.
@end itemize


@node News-3.22.17, News-3.22.16, News-3.22.18, News-3.22.x
@appendixsubsec Changes in release 3.22.17

@itemize @bullet
@item
Fixed a core dump problem when using @code{--log-update} and connecting
without a default database.
@item
Fixed some @code{configure} and portability problems.
@item
Using @code{LEFT JOIN} on tables that had circular dependencies caused
@code{mysqld} to hang forever.
@end itemize


@node News-3.22.16, News-3.22.15, News-3.22.17, News-3.22.x
@appendixsubsec Changes in release 3.22.16 (Feb 1999: Gamma)

@itemize @bullet
@item
@code{mysqladmin processlist} could kill the server if a new user logged in.
@item
@code{DELETE FROM tbl_name WHERE key_column=col_name} didn't find any matching
rows.  Fixed.
@item
@code{DATE_ADD(column, ...)} didn't work.
@item
@code{INSERT DELAYED} could deadlock with status 'upgrading lock'
@item
Extended @code{ENCRYPT()} to take longer salt strings than 2 characters.
@item
@code{longlong2str} is now much faster than before.  For @code{Intel x86}
platforms, this function is written in optimised assembler.
@item
Added the @code{MODIFY} keyword to @code{ALTER TABLE}.
@end itemize


@node News-3.22.15, News-3.22.14, News-3.22.16, News-3.22.x
@appendixsubsec Changes in release 3.22.15

@itemize @bullet
@item
@code{GRANT} used with @code{IDENTIFIED BY} didn't take effect until privileges
were flushed.
@item
Name change of some variables in @code{SHOW STATUS}.
@item
Fixed problem with @code{ORDER BY} with 'only index' optimisation when there
were multiple key definitions for a used column.
@item
@code{DATE} and @code{DATETIME} columns are now up to 5 times faster than
before.
@item
@code{INSERT DELAYED} can be used to let the client do other things while the
server inserts rows into a table.
@item
@code{LEFT JOIN USING (col1,col2)} didn't work if one used it with tables
from 2 different databases.
@item
@code{LOAD DATA LOCAL INFILE} didn't work in the Unix version because of
a missing file.
@item
Fixed problems with @code{VARCHAR}/@code{BLOB} on very short rows (< 4 bytes);
error 127 could occur when deleting rows.
@item
Updating @code{BLOB/TEXT} through formulas didn't work for short (< 256 char)
strings.
@item
When you did a @code{GRANT} on a new host, @code{mysqld} could die on the first
connect from this host.
@item
Fixed bug when one used @code{ORDER BY} on column name that was the same
name as an alias.
@item
Added @code{BENCHMARK(loop_count,expression)} function to time expressions.
@end itemize


@node News-3.22.14, News-3.22.13, News-3.22.15, News-3.22.x
@appendixsubsec Changes in release 3.22.14

@itemize @bullet
@item
Allow empty arguments to @code{mysqld} to make it easier to start from shell
scripts.
@item
Setting a @code{TIMESTAMP} column to @code{NULL} didn't record the timestamp
value in the update log.
@item
Fixed lock handler bug when one did
@code{INSERT INTO TABLE ... SELECT ... GROUP BY}.
@item
Added a patch for @code{localtime_r()} on Windows so that it will not crash
anymore if your date is > 2039, but instead will return a time of all zero.
@item
Names for
user-defined functions are no longer case-sensitive.
@item
Added escape of @code{^Z} (ASCII 26) to @code{\Z} as @code{^Z} doesn't
work with pipes on Windows.
@item
@code{mysql_fix_privileges} adds a new column to the @code{mysql.func} to
support aggregate UDF functions in future MySQL releases.
@end itemize


@node News-3.22.13, News-3.22.12, News-3.22.14, News-3.22.x
@appendixsubsec Changes in release 3.22.13

@itemize @bullet
@item
Saving @code{NOW()}, @code{CURDATE()} or @code{CURTIME()} directly in a
column didn't work.
@item
@code{SELECT COUNT(*) ... LEFT JOIN ...} didn't work with no @code{WHERE} part.
@item
Updated @file{config.guess} to allow MySQL to configure on
UnixWare 7.0.x.
@item
Changed the implementation of @code{pthread_cond()} on the Windows version.
@code{get_lock()} now correctly times out on Windows!
@end itemize


@node News-3.22.12, News-3.22.11, News-3.22.13, News-3.22.x
@appendixsubsec Changes in release 3.22.12

@itemize @bullet
@item
Fixed problem when using @code{DATE_ADD()} and @code{DATE_SUB()} in a
@code{WHERE} clause.
@item
You can now set the password for a user with the @code{GRANT ... TO user
IDENTIFIED BY 'password'} syntax.
@item
Fixed bug in @code{GRANT} checking with @code{SELECT} on many tables.
@item
Added missing file @code{mysql_fix_privilege_tables} to the RPM
distribution. This is not run by default because it relies on the client
package.
@item
Added option @code{SQL_SMALL_RESULT} to @code{SELECT} to force use of
fast temporary tables when you know that the result set will be small.
@item
Allow use of negative real numbers without a decimal point.
@item
Day number is now adjusted to maximum days in month if the resulting month
after @code{DATE_ADD}/@code{DATE_SUB()} doesn't have enough days.
@item
Fix that @code{GRANT} compares columns in case-insensitive fashion.
@item
Fixed a bug in @file{sql_list.h} that made @code{ALTER TABLE} dump core in
some contexts.
@item
The hostname in @code{user@@hostname} can now include @samp{.} and @samp{-}
without quotes in the context of the @code{GRANT}, @code{REVOKE} and
@code{SET PASSWORD FOR ...} statements.
@item
Fix for @code{isamchk} for tables which need big temporary files.
@end itemize


@node News-3.22.11, News-3.22.10, News-3.22.12, News-3.22.x
@appendixsubsec Changes in release 3.22.11

@itemize @bullet
@item
@strong{Important}: You must run the @code{mysql_fix_privilege_tables} script
when you upgrade to this version!  This is needed because of the new
@code{GRANT} system. If you don't do this, you will get @code{Access
denied} when you try to use @code{ALTER TABLE}, @code{CREATE INDEX}, or
@code{DROP INDEX}.
@item
@code{GRANT} to allow/deny users table and column access.
@item
Changed @code{USER()} to return a value in @code{user@@host} format.
Formerly it returned only @code{user}.
@item
Changed the syntax for how to set @code{PASSWORD} for another user.
@item
New command @code{FLUSH STATUS} that resets most status variables to zero.
@item
New status variables: @code{aborted_threads}, @code{aborted_connects}.
@item
New option variable: @code{connection_timeout}.
@item
Added support for Thai sorting (by Pruet Boonma
@email{pruet@@ds90.intanon.nectec.or.th}).
@item
Slovak and japanese error messages.
@item
Configuration and portability fixes.
@item
Added option @code{SET SQL_WARNINGS=1} to get a warning count also for simple
inserts.
@item
MySQL now uses @code{SIGTERM} instead of @code{SIGQUIT} with
shutdown to work better on FreeBSD.
@item
Added option @code{\G} (print vertically) to @code{mysql}.
@item
@code{SELECT HIGH_PRIORITY ...} killed @code{mysqld}.
@item
@code{IS NULL} on a @code{AUTO_INCREMENT} column in a @code{LEFT JOIN} didn't
work as expected.
@item
New function @code{MAKE_SET()}.
@end itemize


@node News-3.22.10, News-3.22.9, News-3.22.11, News-3.22.x
@appendixsubsec Changes in release 3.22.10

@itemize @bullet
@item
@code{mysql_install_db} no longer starts the MySQL server!  You
should start @code{mysqld} with @code{safe_mysqld} after installing it!  The
MySQL RPM will, however, start the server as before.
@item
Added @code{--bootstrap} option to @code{mysqld} and recoded
@code{mysql_install_db} to use it.  This will make it easier to install
MySQL with RPMs.
@item
Changed @code{+}, @code{-} (sign and minus), @code{*}, @code{/}, @code{%},
@code{ABS()} and @code{MOD()} to be @code{BIGINT} aware (64-bit safe).
@item
Fixed a bug in @code{ALTER TABLE} that caused @code{mysqld} to crash.
@item
MySQL now always reports the conflicting key values when a
duplicate key entry occurs. (Before this was only reported for @code{INSERT}.)
@item
New syntax: @code{INSERT INTO tbl_name SET col_name=value, col_name=value, ...}
@item
Most errors in the @file{.err} log are now prefixed with a time stamp.
@item
Added option @code{MYSQL_INIT_COMMAND} to @code{mysql_options()} to make
a query on connect or reconnect.
@item
Added option @code{MYSQL_READ_DEFAULT_FILE} and
@code{MYSQL_READ_DEFAULT_GROUP} to @code{mysql_options()} to read the
following parameters from the MySQL option files: @code{port},
@code{socket}, @code{compress}, @code{password}, @code{pipe}, @code{timeout},
@code{user}, @code{init-command}, @code{host} and @code{database}.
@item
Added @code{maybe_null} to the UDF structure.
@item
Added option @code{IGNORE} to @code{INSERT} statements with many rows.
@item
Fixed some problems with sorting of the @code{koi8} character sets; users of
@code{koi8} @strong{must} run @code{isamchk -rq} on each table that has an
index on a @code{CHAR} or @code{VARCHAR} column.
@item
New script @code{mysql_setpermission}, by Luuk de Boer.  It allows easy
creation of new users with permissions for specific databases.
@item
Allow use of hexadecimal strings (0x...) when specifying a constant string
(like in the column separators with @code{LOAD DATA INFILE}).
@item
Ported to OS/2 (thanks to Antony T. Curtis @email{antony.curtis@@olcs.net}).
@item
Added more variables to @code{SHOW STATUS} and changed format of output to
be like @code{SHOW VARIABLES}.
@item
Added @code{extended-status} command to @code{mysqladmin} which will show the
new status variables.
@end itemize


@node News-3.22.9, News-3.22.8, News-3.22.10, News-3.22.x
@appendixsubsec Changes in release 3.22.9

@itemize @bullet
@item
@code{SET SQL_LOG_UPDATE=0} caused a lockup of the server.
@item
New SQL command:  @code{FLUSH [ TABLES | HOSTS | LOGS | PRIVILEGES ] [, ...]}
@item
New SQL command:  @code{KILL} @code{thread_id}.
@item
Added casts and changed include files to make MySQL easier to
compile on AIX and DEC OSF/1 4.x
@item
Fixed conversion problem when using @code{ALTER TABLE} from a @code{INT}
to a short @code{CHAR()} column.
@item
Added @code{SELECT HIGH_PRIORITY}; this will get a lock for the
@code{SELECT} even if there is a thread waiting for another
@code{SELECT} to get a @code{WRITE LOCK}.
@item
Moved @code{wild_compare()} to string class to be able to use @code{LIKE} on
@code{BLOB}/@code{TEXT} columns with @code{\0}.
@item
Added @code{ESCAPE} option to @code{LIKE}.
@item
Added a lot more output to @code{mysqladmin debug}.
@item
You can now start @code{mysqld} on Windows with the @code{--flush} option.
This will flush all tables to disk after each update.  This makes things
much safer on the Windows platforms but also @strong{much} slower.
@end itemize


@node News-3.22.8, News-3.22.7, News-3.22.9, News-3.22.x
@appendixsubsec Changes in release 3.22.8

@itemize @bullet
@item
Czech character sets should now work much better.  You must also install
@uref{http://www.mysql.com/Downloads/Patches/czech-3.22.8-patch}.
This patch should also be installed if you are using a character set which
uses @code{my_strcoll()}!  The patch should always be safe to install (for any
system), but as this patch changes @code{ISAM} internals it's not yet in the
default distribution.
@item
@code{DATE_ADD()} and @code{DATE_SUB()} didn't work with group functions.
@item
@code{mysql} will now also try to reconnect on @code{USE DATABASE} commands.
@item
Fix problem with @code{ORDER BY} and @code{LEFT JOIN} and @code{const} tables.
@item
Fixed problem with @code{ORDER BY} if the first @code{ORDER BY} column
was a key and the rest of the @code{ORDER BY} columns wasn't part of the key.
@item
Fixed a big problem with @code{OPTIMIZE TABLE}.
@item
MySQL clients on NT will now by default first try to connect with
named pipes and after this with TCP/IP.
@item
Fixed a problem with @code{DROP TABLE} and @code{mysqladmin shutdown} on Windows
(a fatal bug from 3.22.6).
@item
Fixed problems with @code{TIME columns} and negative strings.
@item
Added an extra thread signal loop on shutdown to avoid some error messages
from the client.
@item
MySQL now uses the next available number as extension for the update
log file.
@item
Added patches for UNIXWARE 7.
@end itemize


@node News-3.22.7, News-3.22.6, News-3.22.8, News-3.22.x
@appendixsubsec Changes in release 3.22.7 (Sep 1998: Beta)

@itemize @bullet
@item
Added @code{LIMIT} clause for the @code{DELETE} statement.
@item
You can now use the @code{/*! ... */} syntax to hide MySQL-specific
keywords when you write portable code.  MySQL will parse the code
inside the comments as if the surrounding @code{/*!} and @code{*/} comment
characters didn't exist.
@item
@code{OPTIMIZE TABLE tbl_name} can now be used to reclaim disk space
after many deletes.  Currently, this uses @code{ALTER TABLE} to
regenerate the table, but in the future it will use an integrated
@code{isamchk} for more speed.
@item
Upgraded @code{libtool} to get the configure more portable.
@item
Fixed slow @code{UPDATE} and @code{DELETE} operations when using
@code{DATETIME} or @code{DATE} keys.
@item
Changed optimiser to make it better at deciding when to do a full join
and when using keys.
@item
You can now use @code{mysqladmin proc} to display information about your own
threads. Only users with the @strong{process} privilege can get
information about all threads.
@item
Added handling of formats @code{YYMMDD}, @code{YYYYMMDD},
@code{YYMMDDHHMMSS} for numbers when using @code{DATETIME} and
@code{TIMESTAMP} types. (Formerly these formats only worked with strings.)
@item
Added connect option @code{CLIENT_IGNORE_SPACE} to allow use of spaces
after function names and before @samp{(} (Powerbuilder requires this).
This will make all function names reserved words.
@item
Added the @code{--log-long-format} option to @code{mysqld} to enable timestamps
and INSERT_ID's in the update log.
@item
Added @code{--where} option to @code{mysqldump} (patch by Jim Faucette).
@item
The lexical analyser now uses ``perfect hashing'' for faster parsing of SQL
statements.
@end itemize


@node News-3.22.6, News-3.22.5, News-3.22.7, News-3.22.x
@appendixsubsec Changes in release 3.22.6

@itemize @bullet
@item
Faster @code{mysqldump}.
@item
For the @code{LOAD DATA INFILE} statement, you can now use the new @code{LOCAL}
keyword to read the file from the client.  @code{mysqlimport} will
automatically use @code{LOCAL} when importing with the TCP/IP protocol.
@item
Fixed small optimise problem when updating keys.
@item
Changed makefiles to support shared libraries.
@item
MySQL-NT can now use named pipes, which means that you can now use
MySQL-NT without having to install TCP/IP.
@end itemize


@node News-3.22.5, News-3.22.4, News-3.22.6, News-3.22.x
@appendixsubsec Changes in release 3.22.5

@itemize @bullet
@item
All table lock handing is changed to avoid some very subtle
deadlocks when using @code{DROP TABLE}, @code{ALTER TABLE}, @code{DELETE FROM
TABLE} and @code{mysqladmin flush-tables} under heavy usage.
Changed locking code to get better handling of locks of different types.
@item
Updated @code{DBI} to 1.00 and @code{DBD} to 1.2.0.
@item
Added a check that the error message file contains error messages suitable for
the current version of @code{mysqld}. (To avoid errors if you accidentally
try to use an old error message file.)
@item
All count structures in the client (@code{affected_rows()},
@code{insert_id()}, ...) are now of type @code{BIGINT} to allow 64-bit values
to be used.
This required a minor change in the MySQL protocol which should affect
only old clients when using tables with @code{AUTO_INCREMENT} values > 16M.
@item
The return type of @code{mysql_fetch_lengths()} has changed from @code{uint *}
to @code{ulong *}. This may give a warning for old clients but should work
on most machines.
@item
Change @code{mysys} and @code{dbug} libraries to allocate all thread variables
in one struct.  This makes it easier to make a threaded @file{libmysql.dll}
library.
@item
Use the result from @code{gethostname()} (instead of @code{uname()}) when
constructing @file{.pid} file names.
@item
New better compressed server/client protocol.
@item
@code{COUNT()}, @code{STD()} and @code{AVG()} are extended to handle more than
4G rows.
@item
You can now store values in the range @code{-838:59:59} <= x <=
@code{838:59:59} in a @code{TIME} column.
@item
@strong{Warning: incompatible change!!}
If you set a @code{TIME} column to too short a value, MySQL now
assumes the value is given as: @code{[[[D ]HH:]MM:]SS} instead of
@code{HH[:MM[:SS]]}.
@item
@code{TIME_TO_SEC()} and @code{SEC_TO_TIME()} can now handle negative times
and hours up to 32767.
@item
Added new option
@code{SET OPTION SQL_LOG_UPDATE=@{0|1@}} to allow users with
the @strong{process} privilege to bypass the update log.
(Modified patch from Sergey A Mukhin @email{violet@@rosnet.net}.)
@item
Fixed fatal bug in @code{LPAD()}.
@item
Initialise line buffer in @file{mysql.cc} to make @code{BLOB} reading from
pipes safer.
@item
Added @code{-O max_connect_errors=#} option to @code{mysqld}.
Connect errors are now reset for each correct connection.
@item
Increased the default value of @code{max_allowed_packet} to @code{1M} in
@code{mysqld}.
@item
Added @code{--low-priority-updates} option to @code{mysqld}, to give
table-modifying operations (@code{INSERT}, @code{REPLACE}, @code{UPDATE},
@code{DELETE}) lower priority than retrievals.  You can now use
@code{@{INSERT | REPLACE | UPDATE | DELETE@} LOW_PRIORITY ...} You can
also use @code{SET OPTION SQL_LOW_PRIORITY_UPDATES=@{0|1@}} to change
the priority for one thread.  One side effect is that @code{LOW_PRIORITY}
is now a reserved word. :(
@item
Add support for @code{INSERT INTO table ... VALUES(...),(...),(...)},
to allow inserting multiple rows with a single statement.
@item
@code{INSERT INTO tbl_name} is now also cached when used with @code{LOCK TABLES}.
(Previously only @code{INSERT ... SELECT} and @code{LOAD DATA INFILE} were
cached.)
@item
Allow @code{GROUP BY} functions with @code{HAVING}:
@example
mysql> SELECT col FROM table GROUP BY col HAVING COUNT(*)>0;
@end example
@item
@code{mysqld} will now ignore trailing @samp{;} characters in queries. This
is to make it easier to migrate from some other SQL servers that require the
trailing @samp{;}.
@item
Fix for corrupted fixed-format output generated by @code{SELECT INTO OUTFILE}.
@item
@strong{Warning: incompatible change!}
Added Oracle @code{GREATEST()} and @code{LEAST()} functions.  You must now use
these instead of the @code{MAX()} and @code{MIN()} functions to get the
largest/smallest value from a list of values. These can now handle @code{REAL},
@code{BIGINT} and string (@code{CHAR} or @code{VARCHAR}) values.
@item
@strong{Warning: incompatible change!}
@code{DAYOFWEEK()} had offset 0 for Sunday. Changed the offset to 1.
@item
Give an error for queries that mix @code{GROUP BY} columns and fields when
there is no @code{GROUP BY} specification.
@item
Added @code{--vertical} option to @code{mysql}, for printing results in
vertical mode.
@item
Index-only optimisation; some queries are now resolved using
only indexes.  Until MySQL 4.0, this works only for numeric columns.
@xref{MySQL indexes, , MySQL indexes}.
@item
Lots of new benchmarks.
@item
A new C API chapter and lots of other improvements in the manual.
@end itemize


@node News-3.22.4, News-3.22.3, News-3.22.5, News-3.22.x
@appendixsubsec Changes in release 3.22.4

@itemize @bullet
@item
Added @code{--tmpdir} option to @code{mysqld}, for specifying the location
of the temporary file directory.
@item
MySQL now automatically changes a query from an ODBC client:
@example
SELECT ... FROM table WHERE auto_increment_column IS NULL
@end example
to:
@example
SELECT ... FROM table WHERE auto_increment_column == LAST_INSERT_ID()
@end example
This allows some ODBC programs (Delphi, Access) to retrieve the newly
inserted row to fetch the @code{AUTO_INCREMENT} id.
@item
@code{DROP TABLE} now waits for all users to free a table before deleting it.
@item
Fixed small memory leak in the new connect protocol.
@item
New functions @code{BIN()}, @code{OCT()}, @code{HEX()} and @code{CONV()} for
converting between different number bases.
@item
Added function @code{SUBSTRING()} with 2 arguments.
@item
If you created a table with a record length smaller than 5, you couldn't
delete rows from the table.
@item
Added optimisation to remove const reference tables from @code{ORDER BY} and
@code{GROUP BY}.
@item
@code{mysqld} now automatically disables system locking on Linux and Windows,
and for systems that use MIT-pthreads.  You can force the use of locking
with the @code{--enable-locking} option.
@item
Added @code{--console} option to @code{mysqld}, to force a console window
(for error messages) when using Windows.
@item
Fixed table locks for Windows.
@item
Allow @samp{$} in identifiers.
@item
Changed name of user-specific configuration file from @file{my.cnf} to
@file{.my.cnf} (Unix only).
@item
Added @code{DATE_ADD()} and @code{DATE_SUB()} functions.
@end itemize


@node News-3.22.3, News-3.22.2, News-3.22.4, News-3.22.x
@appendixsubsec Changes in release 3.22.3

@itemize @bullet
@item
Fixed a lock problem (bug in MySQL Version 3.22.1) when closing temporary tables.
@item
Added missing @code{mysql_ping()} to the client library.
@item
Added @code{--compress} option to all MySQL clients.
@item
Changed @code{byte} to @code{char} in @file{mysql.h} and @file{mysql_com.h}.
@end itemize


@node News-3.22.2, News-3.22.1, News-3.22.3, News-3.22.x
@appendixsubsec Changes in release 3.22.2

@itemize @bullet
@item
Searching on multiple constant keys that matched more than 30% of the rows
didn't always use the best possible key.
@item
New functions @code{<<}, @code{>>}, @code{RPAD()} and @code{LPAD()}.
@item
You can now save default options (like passwords) in a configuration file
(@file{my.cnf}).
@item
Lots of small changes to get @code{ORDER BY} to work when no records are found
when using fields that are not in @code{GROUP BY} (MySQL extension).
@item
Added @code{--chroot} option to @code{mysqld}, to start @code{mysqld} in
a chroot environment (by Nikki Chumakov @email{nikkic@@cityline.ru}).
@item
Trailing spaces are now ignored when comparing case-sensitive strings;
this should fix some problems with ODBC and flag 512!
@item
Fixed a core dump bug in the range optimiser.
@item
Added @code{--one-thread} option to @code{mysqld}, for debugging with
LinuxThreads (or @code{glibc}).  (This replaces the @code{-T32} flag)
@item
Added @code{DROP TABLE IF EXISTS} to prevent an error from occurring if the
table doesn't exist.
@item
@code{IF} and @code{EXISTS} are now reserved words (they would have to
be sooner or later).
@item
Added lots of new options to @code{mysqldump}.
@item
Server error messages are now in @file{mysqld_error.h}.
@item
The server/client protocol now supports compression.
@item
All bug fixes from MySQL Version 3.21.32.
@end itemize


@node News-3.22.1, News-3.22.0, News-3.22.2, News-3.22.x
@appendixsubsec Changes in release 3.22.1 (Jun 1998: Alpha)

@itemize @bullet
@item
Added new C API function @code{mysql_ping()}.
@item
Added new API functions @code{mysql_init()} and @code{mysql_options()}.
You now MUST call @code{mysql_init()} before you call
@code{mysql_real_connect()}.
You don't have to call @code{mysql_init()} if you only use
@code{mysql_connect()}.
@item
Added @code{mysql_options(...,MYSQL_OPT_CONNECT_TIMEOUT,...)} so you can set a
timeout for connecting to a server.
@item
Added @code{--timeout} option to @code{mysqladmin}, as a test of
@code{mysql_options()}.
@item
Added @code{AFTER column} and @code{FIRST} options to
@code{ALTER TABLE ... ADD columns}.
This makes it possible to add a new column at some specific location
within a row in an existing table.
@item
@code{WEEK()} now takes an optional argument to allow handling of weeks when
the week starts on Monday (some European countries).  By default,
@code{WEEK()} assumes the week starts on Sunday.
@item
@code{TIME} columns weren't stored properly (bug in MySQL Version 3.22.0).
@item
@code{UPDATE} now returns information about how many rows were
matched and updated, and how many ``warnings'' occurred when doing the update.
@item
Fixed incorrect result from @code{FORMAT(-100,2)}.
@item
@code{ENUM} and @code{SET} columns were compared in binary (case-sensitive)
fashion; changed to be case-insensitive.
@end itemize


@node News-3.22.0,  , News-3.22.1, News-3.22.x
@appendixsubsec Changes in release 3.22.0

@itemize @bullet
@item
New (backward-compatible) connect protocol that allows you to
specify the database to use when connecting, to get much faster
connections to a specific database.

The @code{mysql_real_connect()} call is changed to:

@example
mysql_real_connect(MYSQL *mysql, const char *host, const char *user,
                   const char *passwd, const char *db, uint port,
                   const char *unix_socket, uint client_flag)
@end example
@item
Each connection is handled by its own thread, rather than by the
master @code{accept()} thread.  This fixes permanently the telnet bug
that was a topic on the mail list some time ago.
@item
All TCP/IP connections are now checked with backward-resolution of
the hostname to get better security. @code{mysqld} now has a local hostname
resolver cache so connections should actually be faster than before,
even with this feature.
@item
A site automatically will be blocked from future connections if someone
repeatedly connects with an ``improper header'' (like when one uses telnet).
@item
You can now refer to tables in different databases with references of the form
@code{tbl_name@@db_name} or @code{db_name.tbl_name}.  This makes it possible to
give a user read access to some tables and write access to others simply by
keeping them in different databases!
@item
Added @code{--user} option to @code{mysqld}, to allow it to run
as another Unix user (if it is started as the Unix @code{root} user).
@item
Added caching of users and access rights (for faster access rights checking)
@item
Normal users (not anonymous ones) can change their password with
@code{mysqladmin password 'new_password'}.  This uses encrypted passwords
that are not logged in the normal MySQL log!
@item
All important string functions are now coded in assembler for x86 Linux
machines.  This gives a speedup of 10% in many cases.
@item
For tables that have many columns, the column names are now hashed for
much faster column name lookup (this will speed up some benchmark
tests a lot!)
@item
Some benchmarks are changed to get better individual timing.
(Some loops were so short that a specific test took < 2 seconds. The
loops have been changed to take about 20 seconds to make it easier
to compare different databases. A test that took 1-2 seconds before
now takes 11-24 seconds, which is much better)
@item
Re-arranged @code{SELECT} code to handle some very specific queries
involving group functions (like @code{COUNT(*)}) without a @code{GROUP BY} but
with @code{HAVING}. The following now works:
@example
mysql> SELECT COUNT(*) as C FROM table HAVING C > 1;
@end example
@item
Changed the protocol for field functions to be faster and avoid some
calls to @code{malloc()}.
@item
Added @code{-T32} option to @code{mysqld}, for running all queries under the
main thread.  This makes it possible to debug @code{mysqld} under Linux with
@code{gdb}!
@item
Added optimisation of @code{not_null_column IS NULL} (needed for some Access
queries).
@item
Allow @code{STRAIGHT_JOIN} to be used between two tables to force the optimiser
to join them in a specific order.
@item
String functions now return @code{VARCHAR} rather than @code{CHAR} and
the column type is now @code{VARCHAR} for fields saved as @code{VARCHAR}.
This should make the @code{MyODBC} driver better, but may break some old
MySQL clients that don't handle @code{FIELD_TYPE_VARCHAR} the same
way as @code{FIELD_TYPE_CHAR}.
@item
@code{CREATE INDEX} and @code{DROP INDEX} are now implemented through
@code{ALTER TABLE}.
@code{CREATE TABLE} is still the recommended (fast) way to create indexes.
@item
Added @code{--set-variable} option @code{wait_timeout} to @code{mysqld}.
@item
Added time column to @code{mysqladmin processlist} to show how long a query
has taken or how long a thread has slept.
@item
Added lots of new variables to @code{show variables} and some new to
@code{show status}.
@item
Added new type @code{YEAR}. @code{YEAR} is stored in 1 byte with allowable
values of 0, and 1901 to 2155.
@item
Added new @code{DATE} type that is stored in 3 bytes rather than 4 bytes.
All new tables are created with the new date type if you don't use the
@code{--old-protocol} option to @code{mysqld}.
@item
Fixed bug in record caches; for some queries, you could get
@code{Error from table handler: #} on some operating systems.
@item
Added @code{--enable-assembler} option to @code{configure}, for x86 machines
(tested on Linux + @code{gcc}).  This will enable assembler functions for the
most important string functions for more speed!
@end itemize


@node News-3.21.x, News-3.20.x, News-3.22.x, News
@appendixsec Changes in release 3.21.x

@cindex changes, version 3.21

Version 3.21 is quite old now, and should be avoided if possible.  This
information is kept here for historical purposes only.

@menu
* News-3.21.33::                Changes in release 3.21.33
* News-3.21.32::                Changes in release 3.21.32
* News-3.21.31::                Changes in release 3.21.31
* News-3.21.30::                Changes in release 3.21.30
* News-3.21.29::                Changes in release 3.21.29
* News-3.21.28::                Changes in release 3.21.28
* News-3.21.27::                Changes in release 3.21.27
* News-3.21.26::                Changes in release 3.21.26
* News-3.21.25::                Changes in release 3.21.25
* News-3.21.24::                Changes in release 3.21.24
* News-3.21.23::                Changes in release 3.21.23
* News-3.21.22::                Changes in release 3.21.22
* News-3.21.21a::               Changes in release 3.21.21a
* News-3.21.21::                Changes in release 3.21.21
* News-3.21.20::                Changes in release 3.21.20
* News-3.21.19::                Changes in release 3.21.19
* News-3.21.18::                Changes in release 3.21.18
* News-3.21.17::                Changes in release 3.21.17
* News-3.21.16::                Changes in release 3.21.16
* News-3.21.15::                Changes in release 3.21.15
* News-3.21.14b::               Changes in release 3.21.14b
* News-3.21.14a::               Changes in release 3.21.14a
* News-3.21.13::                Changes in release 3.21.13
* News-3.21.12::                Changes in release 3.21.12
* News-3.21.11::                Changes in release 3.21.11
* News-3.21.10::                Changes in release 3.21.10
* News-3.21.9::                 Changes in release 3.21.9
* News-3.21.8::                 Changes in release 3.21.8
* News-3.21.7::                 Changes in release 3.21.7
* News-3.21.6::                 Changes in release 3.21.6
* News-3.21.5::                 Changes in release 3.21.5
* News-3.21.4::                 Changes in release 3.21.4
* News-3.21.3::                 Changes in release 3.21.3
* News-3.21.2::                 Changes in release 3.21.2
* News-3.21.0::                 Changes in release 3.21.0
@end menu


@node News-3.21.33, News-3.21.32, News-3.21.x, News-3.21.x
@appendixsubsec Changes in release 3.21.33

@itemize @bullet
@item
Fixed problem when sending @code{SIGHUP} to @code{mysqld};
@code{mysqld} core dumped when starting from boot on some systems.
@item
Fixed problem with losing a little memory for some connections.
@item
@code{DELETE FROM tbl_name} without a @code{WHERE} condition is now done the
long way when you use @code{LOCK TABLES} or if the table is in use, to
avoid race conditions.
@item
@code{INSERT INTO TABLE (timestamp_column) VALUES (NULL);} didn't set timestamp.
@end itemize


@node News-3.21.32, News-3.21.31, News-3.21.33, News-3.21.x
@appendixsubsec Changes in release 3.21.32

@itemize @bullet
@item
Fixed some possible race conditions when doing many reopen/close on the same
tables under heavy load!  This can happen if you execute @code{mysqladmin
refresh} often. This could in some very rare cases corrupt the header of the
index file and cause error 126 or 138.

@item
Fixed fatal bug in @code{refresh()} when running with the
@code{--skip-locking} option.  There was a ``very small'' time gap after
a @code{mysqladmin refresh} when a table could be corrupted if one
thread updated a table while another thread did @code{mysqladmin
refresh} and another thread started a new update ont the same table
before the first thread had finished.  A refresh (or
@code{--flush-tables}) will now not return until all used tables are
closed!
@item
@code{SELECT DISTINCT} with a @code{WHERE} clause that didn't match any rows
returned a row in some contexts (bug only in 3.21.31).
@item
@code{GROUP BY} + @code{ORDER BY} returned one empty row when no rows where
found.
@item
Fixed a bug in the range optimiser that wrote
@code{Use_count: Wrong count for ...} in the error log file.
@end itemize


@node News-3.21.31, News-3.21.30, News-3.21.32, News-3.21.x
@appendixsubsec Changes in release 3.21.31

@itemize @bullet
@item
Fixed a sign extension problem for the @code{TINYINT} type on Irix.
@item
Fixed problem with @code{LEFT("constant_string",function)}.
@item
Fixed problem with @code{FIND_IN_SET()}.
@item
@code{LEFT JOIN} core dumped if the second table is used with a constant
@code{WHERE/ON} expression that uniquely identifies one record.
@item
Fixed problems with @code{DATE_FORMAT()} and incorrect dates.
@code{DATE_FORMAT()} now ignores @code{'%'} to make it possible to extend
it more easily in the future.
@end itemize


@node News-3.21.30, News-3.21.29, News-3.21.31, News-3.21.x
@appendixsubsec Changes in release 3.21.30

@itemize @bullet
@item
@code{mysql} now returns an exit code > 0 if the query returned an error.
@item
Saving of command-line history to file in @code{mysql} client.
By Tommy Larsen @email{tommy@@mix.hive.no}.
@item
Fixed problem with empty lines that were ignored in @file{mysql.cc}.
@item
Save the pid of the signal handler thread in the pid file instead
of the pid of the main thread.
@item
Added patch by @email{tommy@@valley.ne.jp} to support Japanese characters
SJIS and UJIS.
@item
Changed @code{safe_mysqld} to redirect startup messages to
@code{'hostname'.err} instead
of @code{'hostname'.log} to reclaim file space on @code{mysqladmin refresh}.
@item
@code{ENUM} always had the first entry as default value.
@item
@code{ALTER TABLE} wrote two entries to the update log.
@item
@code{sql_acc()} now closes the @code{mysql} grant tables after a reload to
save table space and memory.
@item
Changed @code{LOAD DATA} to use less memory with tables and @code{BLOB}
columns.
@item
Sorting on a function which made a division / 0 produced a wrong set
in some cases.
@item
Fixed @code{SELECT} problem with @code{LEFT()} when using the @code{czech}
character set.
@item
Fixed problem in @code{isamchk}; it couldn't repair a packed table in a very
unusual case.
@item
@code{SELECT} statements with @code{&} or @code{|} (bit functions) failed on
columns with @code{NULL} values.
@item
When comparing a field = field, where one of the fields was a part key,
only the length of the part key was compared.
@end itemize


@node News-3.21.29, News-3.21.28, News-3.21.30, News-3.21.x
@appendixsubsec Changes in release 3.21.29

@itemize @bullet
@item
@code{LOCK TABLES} + @code{DELETE from tbl_name} never removed locks properly.
@item
Fixed problem when grouping on an @code{OR} function.
@item
Fixed permission problem with @code{umask()} and creating new databases.
@item
Fixed permission problem on result file with @code{SELECT ... INTO OUTFILE ...}
@item
Fixed problem in range optimiser (core dump) for a very complex query.
@item
Fixed problem when using @code{MIN(integer)} or @code{MAX(integer)} in
@code{GROUP BY}.
@item
Fixed bug on Alpha when using integer keys. (Other keys worked on Alpha.)
@item
Fixed bug in @code{WEEK("XXXX-xx-01")}.
@end itemize


@node News-3.21.28, News-3.21.27, News-3.21.29, News-3.21.x
@appendixsubsec Changes in release 3.21.28

@itemize @bullet
@item
Fixed socket permission (clients couldn't connect to Unix socket on Linux).
@item
Fixed bug in record caches; for some queries, you could get
@code{Error from table handler: #} on some operating systems.
@end itemize


@node News-3.21.27, News-3.21.26, News-3.21.28, News-3.21.x
@appendixsubsec Changes in release 3.21.27

@itemize @bullet
@item
Added user level lock functions @code{GET_LOCK(string,timeout)},
@code{RELEASE_LOCK(string)}.
@item
Added @code{Opened_tables} to @code{show status}.
@item
Changed connect timeout to 3 seconds to make it somewhat harder
for crackers to kill @code{mysqld} through telnet + TCP/IP.
@item
Fixed bug in range optimiser when using
@code{WHERE key_part_1 >= something AND key_part_2 <= something_else}.
@item
Changed @code{configure} for detection of FreeBSD 3.0 9803xx and above
@item
@code{WHERE} with @code{string_col_key = constant_string} didn't always
find all rows if the column had many values differing only with
characters of the same sort value (like e and @'e).
@item
Strings keys looked up with 'ref' were not compared in case-sensitive fashion.
@item
Added @code{umask()} to make log files non-readable for normal users.
@item
Ignore users with old (8-byte) password on startup if not using
@code{--old-protocol} option to @code{mysqld}.
@item
@code{SELECT} which matched all key fields returned the values in the
case of the matched values, not of the found values. (Minor problem.)
@end itemize


@node News-3.21.26, News-3.21.25, News-3.21.27, News-3.21.x
@appendixsubsec Changes in release 3.21.26

@itemize @bullet
@item
@code{FROM_DAYS(0)} now returns "0000-00-00".
@item
In @code{DATE_FORMAT()}, PM and AM were swapped for hours 00 and 12.
@item
Extended the default maximum key size to 256.
@item
Fixed bug when using @code{BLOB}/@code{TEXT} in @code{GROUP BY} with many
tables.
@item
An @code{ENUM} field that is not declared @code{NOT NULL} has @code{NULL} as
the default value.
(Previously, the default value was the first enumeration value.)
@item
Fixed bug in the join optimiser code when using many part keys
on the same key:  @code{INDEX (Organisation,Surname(35),Initials(35))}.
@item
Added some tests to the table order optimiser to get some cases with
@code{SELECT ... FROM many_tables} much faster.
@item
Added a retry loop around @code{accept()} to possibly fix some problems on some
Linux machines.
@end itemize


@node News-3.21.25, News-3.21.24, News-3.21.26, News-3.21.x
@appendixsubsec Changes in release 3.21.25

@itemize @bullet
@item
Changed @code{typedef 'string'} to @code{typedef 'my_string'} for better
portability.
@item
You can now kill threads that are waiting on a disk-full condition.
@item
Fixed some problems with UDF functions.
@item
Added long options to @code{isamchk}. Try @code{isamchk --help}.
@item
Fixed a bug when using 8 bytes long (alpha); @code{filesort()} didn't work.
Affects @code{DISTINCT}, @code{ORDER BY} and @code{GROUP BY} on 64-bit
processors.
@end itemize


@node News-3.21.24, News-3.21.23, News-3.21.25, News-3.21.x
@appendixsubsec Changes in release 3.21.24

@itemize @bullet
@item
Dynamic loadable functions. Based on source from Alexis Mikhailov.
@item
You couldn't delete from a table if no one had done a @code{SELECT} on the
table.
@item
Fixed problem with range optimiser with many @code{OR} operators on key parts
inside each other.
@item
Recoded @code{MIN()} and @code{MAX()} to work properly with strings and
@code{HAVING}.
@item
Changed default umask value for new files from @code{0664} to @code{0660}.
@item
Fixed problem with @code{LEFT JOIN} and constant expressions in the @code{ON}
part.
@item
Added Italian error messages from @email{brenno@@dewinter.com}.
@item
@code{configure} now works better on OSF/1 (tested on 4.0D).
@item
Added hooks to allow @code{LIKE} optimisation with international character
support.
@item
Upgraded @code{DBI} to 0.93.
@end itemize


@node News-3.21.23, News-3.21.22, News-3.21.24, News-3.21.x
@appendixsubsec Changes in release 3.21.23

@itemize @bullet
@item
The following symbols are now reserved words:
@code{TIME}, @code{DATE}, @code{TIMESTAMP}, @code{TEXT}, @code{BIT},
@code{ENUM}, @code{NO}, @code{ACTION}, @code{CHECK}, @code{YEAR},
@code{MONTH}, @code{DAY}, @code{HOUR}, @code{MINUTE}, @code{SECOND},
@code{STATUS}, @code{VARIABLES}.
@item
Setting a @code{TIMESTAMP} to @code{NULL} in @code{LOAD DATA INFILE ...} didn't
set the current time for the @code{TIMESTAMP}.
@item
Fix @code{BETWEEN} to recognise binary strings. Now @code{BETWEEN} is
case-sensitive.
@item
Added @code{--skip-thread-priority} option to @code{mysqld}, for systems
where @code{mysqld}'s thread scheduling doesn't work properly (BSDI 3.1).
@item
Added ODBC functions @code{DAYNAME()} and @code{MONTHNAME()}.
@item
Added function @code{TIME_FORMAT()}.  This works like @code{DATE_FORMAT()},
but takes a time string (@code{'HH:MM:DD'}) as argument.
@item
Fixed unlikely(?) key optimiser bug when using @code{OR}s of key parts
inside @code{AND}s.
@item
Added @code{variables} command to @code{mysqladmin}.
@item
A lot of small changes to the binary releases.
@item
Fixed a bug in the new protocol from MySQL Version 3.21.20.
@item
Changed @code{ALTER TABLE} to work with Windows (Windows can't rename
open files).  Also fixed a couple of small bugs in the Windows version.
@item
All standard MySQL clients are now ported to MySQL-Windows.
@item
MySQL can now be started as a service on NT.
@end itemize


@node News-3.21.22, News-3.21.21a, News-3.21.23, News-3.21.x
@appendixsubsec Changes in release 3.21.22

@itemize @bullet
@item
Starting with this version, all MySQL distributions will be
configured, compiled and tested with @code{crash-me} and the benchmarks on
the following platforms: SunOS 5.6 sun4u, SunOS 5.5.1 sun4u, SunOS 4.14 sun4c,
SunOS 5.6 i86pc, Irix 6.3 mips5k, HP-UX 10.20 hppa, AIX 4.2.1 ppc,
OSF/1 V4.0 alpha, FreeBSD 2.2.2 i86pc and BSDI 3.1 i386.
@item
Fix @code{COUNT(*)} problems when the @code{WHERE} clause didn't match any
records.  (Bug from 3.21.17.)
@item
Removed that @code{NULL = NULL} is true.  Now you must use @code{IS NULL}
or @code{IS NOT NULL} to test whether a value is @code{NULL}.
(This is according to ANSI SQL but may break
old applications that are ported from @code{mSQL}.)
You can get the old behavior by compiling with @code{-DmSQL_COMPLIANT}.
@item
Fixed bug that core dumped when using many @code{LEFT OUTER JOIN} clauses.
@item
Fixed bug in @code{ORDER BY} on string formula with possible @code{NULL} values.
@item
Fixed problem in range optimiser when using @code{<=} on sub index.
@item
Added functions @code{DAYOFYEAR()}, @code{DAYOFMONTH()}, @code{MONTH()},
@code{YEAR()}, @code{WEEK()}, @code{QUARTER()}, @code{HOUR()}, @code{MINUTE()},
@code{SECOND()} and @code{FIND_IN_SET()}.
@item
Added @code{SHOW VARIABLES} command.
@item
Added support of ``long constant strings'' from ANSI SQL:
@example
mysql> SELECT 'first ' 'second';       -> 'first second'
@end example
@item
Upgraded Msql-Mysql-modules to 1.1825.
@item
Upgraded @code{mysqlaccess} to 2.02.
@item
Fixed problem with Russian character set and @code{LIKE}.
@item
Ported to OpenBSD 2.1.
@item
New Dutch error messages.
@end itemize


@node News-3.21.21a, News-3.21.21, News-3.21.22, News-3.21.x
@appendixsubsec Changes in release 3.21.21a

@itemize @bullet
@item
Configure changes for some operating systems.
@end itemize


@node News-3.21.21, News-3.21.20, News-3.21.21a, News-3.21.x
@appendixsubsec Changes in release 3.21.21

@itemize @bullet
@item
Fixed optimiser bug when using
@code{WHERE data_field = date_field2 AND date_field2 = constant}.
@item
Added @code{SHOW STATUS} command.
@item
Removed @file{manual.ps} from the source distribution to make it smaller.
@end itemize


@node News-3.21.20, News-3.21.19, News-3.21.21, News-3.21.x
@appendixsubsec Changes in release 3.21.20

@itemize @bullet
@item
Changed the maximum table name and column name lengths from 32 to 64.
@item
Aliases can now be of ``any'' length.
@item
Fixed @code{mysqladmin stat} to return the right number of queries.
@item
Changed protocol (downward compatible) to mark if a column has the
@code{AUTO_INCREMENT} attribute or is a @code{TIMESTAMP}. This is needed for
the new Java driver.
@item
Added Hebrew sorting order by Zeev Suraski.
@item
Solaris 2.6: Fixed @code{configure} bugs and increased maximum table size
from 2G to 4G.
@end itemize


@node News-3.21.19, News-3.21.18, News-3.21.20, News-3.21.x
@appendixsubsec Changes in release 3.21.19

@itemize @bullet
@item
Upgraded @code{DBD} to 1.1823. This version implements @code{mysql_use_result}
in @code{DBD-Mysql}.
@item
Benchmarks updated for empress (by Luuk).
@item
Fixed a case of slow range searching.
@item
Configure fixes (@file{Docs} directory).
@item
Added function @code{REVERSE()} (by Zeev Suraski).
@end itemize


@node News-3.21.18, News-3.21.17, News-3.21.19, News-3.21.x
@appendixsubsec Changes in release 3.21.18

@itemize @bullet
@item
Issue error message if client C functions are called in wrong order.
@item
Added automatic reconnect to the @file{libmysql.c} library. If a write command
fails, an automatic reconnect is done.
@item
Small sort sets no longer use temporary files.
@item
Upgraded @code{DBI} to 0.91.
@item
Fixed a couple of problems with @code{LEFT OUTER JOIN}.
@item
Added @code{CROSS JOIN} syntax. @code{CROSS} is now a reserved word.
@item
Recoded @code{yacc}/@code{bison} stack allocation to be even safer and to allow
MySQL to handle even bigger expressions.
@item
Fixed a couple of problems with the update log.
@item
@code{ORDER BY} was slow when used with key ranges.
@end itemize


@node News-3.21.17, News-3.21.16, News-3.21.18, News-3.21.x
@appendixsubsec Changes in release 3.21.17

@itemize @bullet
@item
Changed documentation string of @code{--with-unix-socket-path} to avoid
confusion.
@item
Added ODBC and ANSI SQL style @code{LEFT OUTER JOIN}.
@item
The following are new reserved words: @code{LEFT}, @code{NATURAL},
@code{USING}.
@item
The client library now uses the value of the environment variable
@code{MYSQL_HOST} as the default host if it's defined.
@item
@code{SELECT col_name, SUM(expr)} now returns @code{NULL} for @code{col_name}
when there are matching rows.
@item
Fixed problem with comparing binary strings and @code{BLOB}s with ASCII
characters over 127.
@item
Fixed lock problem: when freeing a read lock on a table with multiple
read locks, a thread waiting for a write lock would have been given the lock.
This shouldn't affect data integrity, but could possibly make @code{mysqld}
restart if one thread was reading data that another thread modified.
@item
@code{LIMIT offset,count} didn't work in @code{INSERT ... SELECT}.
@item
Optimised key block caching. This will be quicker than the old algorithm when
using bigger key caches.
@end itemize


@node News-3.21.16, News-3.21.15, News-3.21.17, News-3.21.x
@appendixsubsec Changes in release 3.21.16

@itemize @bullet
@item
Added ODBC 2.0 & 3.0 functions @code{POWER()}, @code{SPACE()},
@code{COT()}, @code{DEGREES()}, @code{RADIANS()}, @code{ROUND(2 arg)}
and @code{TRUNCATE()}.
@item
@strong{Warning: Incompatible change!} @code{LOCATE()} parameters were
swapped according to ODBC standard. Fixed.
@item
Added function @code{TIME_TO_SEC()}.
@item
In some cases, default values were not used for @code{NOT NULL} fields.
@item
Timestamp wasn't always updated properly in @code{UPDATE SET ...} statements.
@item
Allow empty strings as default values for @code{BLOB} and @code{TEXT}, to
be compatible with @code{mysqldump}.
@end itemize


@node News-3.21.15, News-3.21.14b, News-3.21.16, News-3.21.x
@appendixsubsec Changes in release 3.21.15

@itemize @bullet
@item
@strong{Warning: Incompatible change!} @code{mysqlperl} is now from
Msql-Mysql-modules. This means that @code{connect()} now takes
@code{host}, @code{database}, @code{user}, @code{password} arguments! The old
version took @code{host}, @code{database}, @code{password}, @code{user}.
@item
Allow @code{DATE '1997-01-01'}, @code{TIME '12:10:10'} and
@code{TIMESTAMP '1997-01-01 12:10:10'} formats required by ANSI SQL.
@strong{Warning: Incompatible change!} This has the unfortunate
side-effect that you no longer can have columns named @code{DATE}, @code{TIME}
or @code{TIMESTAMP}. :( Old columns can still be accessed through
@code{tablename.columnname}!)
@item
Changed Makefiles to hopefully work better with BSD systems. Also,
@file{manual.dvi} is now included in the distribution to avoid having stupid
@code{make} programs trying to rebuild it.
@item
@code{readline} library upgraded to version 2.1.
@item
A new sortorder german-1. That is a normal ISO-Latin1 with a german sort
order.
@item
Perl @code{DBI}/@code{DBD} is now included in the distribution. @code{DBI}
is now the recommended way to connect to MySQL from Perl.
@item
New portable benchmark suite with @code{DBD}, with test results from
@code{mSQL} 2.0.3, MySQL, PostgreSQL 6.2.1 and Solid server 2.2.
@item
@code{crash-me} is now included with the benchmarks; this is a Perl program
designed to find as many limits as possible in a SQL server. Tested with
@code{mSQL}, PostgreSQL, Solid and MySQL.
@item
Fixed bug in range-optimiser that crashed MySQL on some queries.
@item
Table and column name completion for @code{mysql} command-line tool, by Zeev
Suraski and Andi Gutmans.
@item
Added new command @code{REPLACE} that works like @code{INSERT} but
replaces conflicting records with the new record. @code{REPLACE INTO
TABLE ... SELECT ...} works also.
@item
Added new commands @code{CREATE DATABASE db_name} and @code{DROP
DATABASE db_name}.
@item
Added @code{RENAME} option to @code{ALTER TABLE}: @code{ALTER TABLE name
RENAME TO new_name}.
@item
@code{make_binary_distribution} now includes @file{libgcc.a} in
@file{libmysqlclient.a}. This should make linking work for people who don't
have @code{gcc}.
@item
Changed @code{net_write()} to @code{my_net_write()} because of a name
conflict with Sybase.
@item
@cindex ODBC compatibility
@cindex compatibility, with ODBC
New function @code{DAYOFWEEK()} compatible with ODBC.
@item
Stack checking and @code{bison} memory overrun checking to make MySQL
safer with weird queries.
@end itemize


@node News-3.21.14b, News-3.21.14a, News-3.21.15, News-3.21.x
@appendixsubsec Changes in release 3.21.14b

@itemize @bullet
@item
Fixed a couple of small @code{configure} problems on some platforms.
@end itemize


@node News-3.21.14a, News-3.21.13, News-3.21.14b, News-3.21.x
@appendixsubsec Changes in release 3.21.14a

@itemize @bullet
@item
Ported to SCO Openserver 5.0.4  with FSU Pthreads.
@item
HP-UX 10.20 should work.
@item
Added new function @code{DATE_FORMAT()}.
@item
Added @code{NOT IN}.
@item
Added automatic removal of 'ODBC function conversions': @code{@{fn now() @}}
@item
Handle ODBC 2.50.3 option flags.
@item
Fixed comparison of @code{DATE} and @code{TIME} values with @code{NULL}.
@item
Changed language name from germany to german to be consistent with
the other language names.
@item
Fixed sorting problem on functions returning a @code{FLOAT}. Previously, the
values were converted to @code{INT}s before sorting.
@item
Fixed slow sorting when sorting on key field when using
@code{key_column=constant}.
@item
Sorting on calculated @code{DOUBLE} values sorted on integer results instead.
@item
@code{mysql} no longer requires a database argument.
@item
Changed the place where @code{HAVING} should be. According to ANSI, it should
be after @code{GROUP BY} but before @code{ORDER BY}. MySQL Version 3.20
incorrectly had it last.
@item
Added Sybase command @code{USE DATABASE} to start using another database.
@item
Added automatic adjusting of number of connections and table cache size
if the maximum number of files that can be opened is less than needed.
This should fix that @code{mysqld} doesn't crash even if you haven't done a
@code{ulimit -n 256} before starting @code{mysqld}.
@item
Added lots of limit checks to make it safer when running with too little
memory or when doing weird queries.
@end itemize


@node News-3.21.13, News-3.21.12, News-3.21.14a, News-3.21.x
@appendixsubsec Changes in release 3.21.13

@itemize @bullet
@item
Added retry of interrupted reads and clearing of @code{errno}.
This makes Linux systems much safer!
@item
Fixed locking bug when using many aliases on the same table in the same
@code{SELECT}.
@item
Fixed bug with @code{LIKE} on number key.
@item
New error message so you can check whether the connection was lost while
the command was running or whether the connection was down from the start.
@item
Added @code{--table} option to @code{mysql} to print in table format.
Moved time and row information after query result.
Added automatic reconnect of lost connections.
@item
Added @code{!=} as a synonym for @code{<>}.
@item
Added function @code{VERSION()} to make easier logs.
@item
New multi-user test @file{tests/fork_test.pl} to put some strain on the
thread library.
@end itemize


@node News-3.21.12, News-3.21.11, News-3.21.13, News-3.21.x
@appendixsubsec Changes in release 3.21.12

@itemize @bullet
@item
Fixed @code{ftruncate()} call in MIT-pthreads. This made @code{isamchk}
destroy the @file{.ISM} files on (Free)BSD 2.x systems.
@item
Fixed broken @code{__P_} patch in MIT-pthreads.
@item
Many memory overrun checks. All string functions now return @code{NULL}
if the returned string should be longer than @code{max_allowed_packet} bytes.
@item
Changed the name of the @code{INTERVAL} type to @code{ENUM}, because
@code{INTERVAL} is used in ANSI SQL.
@item
In some cases, doing a @code{JOIN} + @code{GROUP} + @code{INTO OUTFILE},
the result wasn't grouped.
@item
@code{LIKE} with @code{'_'} as last character didn't work. Fixed.
@item
Added extended ANSI SQL @code{TRIM()} function.
@item
Added @code{CURTIME()}.
@item
Added @code{ENCRYPT()} function by Zeev Suraski.
@item
Fixed better @code{FOREIGN KEY} syntax skipping. New reserved words:
@code{MATCH}, @code{FULL}, @code{PARTIAL}.
@item
@code{mysqld} now allows IP number and hostname for the @code{--bind-address}
option.
@item
Added @code{SET OPTION CHARACTER SET cp1251_koi8} to enable conversions of
data to and from the @code{cp1251_koi8} character set.
@item
Lots of changes for Windows 95 port. In theory, this version should now be
easily portable to Windows 95.
@item
Changed the @code{CREATE COLUMN} syntax of @code{NOT NULL} columns to be after
the @code{DEFAULT} value, as specified in the ANSI SQL standard. This will
make @code{mysqldump} with @code{NOT NULL} and default values incompatible with
MySQL Version 3.20.
@item
Added many function name aliases so the functions can be used with
ODBC or ANSI SQL92 syntax.
@item
Fixed syntax of @code{ALTER TABLE tbl_name ALTER COLUMN col_name SET DEFAULT
NULL}.
@item
Added @code{CHAR} and @code{BIT} as synonyms for @code{CHAR(1)}.
@item
Fixed core dump when updating as a user who has only @strong{select} privilege.
@item
@code{INSERT ... SELECT ... GROUP BY} didn't work in some cases. An
@code{Invalid use of group function} error occurred.
@item
When using @code{LIMIT}, @code{SELECT} now always uses keys instead of record
scan.  This will give better performance on @code{SELECT} and a @code{WHERE}
that matches many rows.
@item
Added Russian error messages.
@end itemize


@node News-3.21.11, News-3.21.10, News-3.21.12, News-3.21.x
@appendixsubsec Changes in release 3.21.11

@itemize @bullet
@item
Configure changes.
@item
MySQL now works with the new thread library on BSD/OS 3.0.
@item
Added new group functions @code{BIT_OR()} and @code{BIT_AND()}.
@item
Added compatibility functions @code{CHECK} and @code{REFERENCES}.
@code{CHECK} is now a reserved word.
@item
Added @code{ALL} option to @code{GRANT} for better compatibility. (@code{GRANT}
is still a dummy function.)
@item
Added partly-translated dutch messages.
@item
Fixed bug in @code{ORDER BY} and @code{GROUP BY} with @code{NULL} columns.
@item
Added function @code{last_insert_id()} to retrieve last @code{AUTO_INCREMENT}
value.  This is intended for clients to ODBC that can't use the
@code{mysql_insert_id()} API function, but can be used by any client.
@item
Added @code{--flush-logs} option to @code{mysqladmin}.
@item
Added command @code{STATUS} to @code{mysql}.
@item
Fixed problem with @code{ORDER BY}/@code{GROUP BY} because of bug in @code{gcc}.
@item
Fixed problem with @code{INSERT ... SELECT ... GROUP BY}.
@end itemize


@node News-3.21.10, News-3.21.9, News-3.21.11, News-3.21.x
@appendixsubsec Changes in release 3.21.10

@itemize @bullet
@item
New @code{mysqlaccess}.
@item
@code{CREATE} now supports all ODBC types and the @code{mSQL} @code{TEXT} type.
All ODBC 2.5 functions are also supported (added @code{REPEAT}). This provides
better portability.
@item
Added text types @code{TINYTEXT}, @code{TEXT}, @code{MEDIUMTEXT} and
@code{LONGTEXT}.  These are actually @code{BLOB}types, but all searching is
done in case-insensitive fashion.
@item
All old @code{BLOB} fields are now @code{TEXT} fields.  This only
changes that all searching on strings is done in case-sensitive fashion.
You must do an @code{ALTER TABLE} and change the field type to @code{BLOB}
if you want to have tests done in case-sensitive fashion.
@item
Fixed some @code{configure} issues.
@item
Made the locking code a bit safer. Fixed very unlikely
deadlock situation.
@item
Fixed a couple of bugs in the range optimiser. Now the new range benchmark
@code{test-select} works.
@end itemize


@node News-3.21.9, News-3.21.8, News-3.21.10, News-3.21.x
@appendixsubsec Changes in release 3.21.9

@itemize @bullet
@item
Added @code{--enable-unix-socket=pathname} option to @code{configure}.
@item
Fixed a couple of portability problems with include files.
@item
Fixed bug in range calculation that could return empty
set when searching on multiple key with only one entry (very rare).
@item
Most things ported to FSU Pthreads, which should allow MySQL to
run on Caldera (SCO). @xref{Caldera}.
@end itemize


@node News-3.21.8, News-3.21.7, News-3.21.9, News-3.21.x
@appendixsubsec Changes in release 3.21.8

@itemize @bullet
@item
Works now in Solaris 2.6.
@item
Added handling of calculation of @code{SUM()} functions.
For example, you can now use @code{SUM(column)/COUNT(column)}.
@item
Added handling of trigometric functions:
@code{PI()}, @code{ACOS()}, @code{ASIN()}, @code{ATAN()}, @code{COS()},
@code{SIN()} and @code{TAN()}.
@item
New languages: norwegian, norwegian-ny and portuguese.
@item
Fixed parameter bug in @code{net_print()} in @file{procedure.cc}.
@item
Fixed a couple of memory leaks.
@item
Now allow also the old @code{SELECT ... INTO OUTFILE} syntax.
@item
Fixed bug with @code{GROUP BY} and @code{SELECT} on key with many values.
@item
@code{mysql_fetch_lengths()} sometimes returned incorrect lengths when you used
@code{mysql_use_result()}. This affected at least some cases of
@code{mysqldump --quick}.
@item
Fixed bug in optimisation of @code{WHERE const op field}.
@item
Fixed problem when sorting on @code{NULL} fields.
@item
Fixed a couple of 64-bit (Alpha) problems.
@item
Added @code{--pid-file=#} option to @code{mysqld}.
@item
Added date formatting to @code{FROM_UNIXTIME()}, originally by Zeev Suraski.
@item
Fixed bug in @code{BETWEEN} in range optimiser (did only test = of the first
argument).
@item
Added machine-dependent files for MIT-pthreads i386-SCO. There is probably
more to do to get this to work on SCO 3.5.
@end itemize


@node News-3.21.7, News-3.21.6, News-3.21.8, News-3.21.x
@appendixsubsec Changes in release 3.21.7

@itemize @bullet
@item
Changed @file{Makefile.am} to take advantage of Automake 1.2.
@item
Added the beginnings of a benchmark suite.
@item
Added more secure password handling.
@item
Added new client function @code{mysql_errno()}, to get the error number of
the error message. This makes error checking in the client much easier.
This makes the new server incompatible with the 3.20.x server when running
without @code{--old-protocol}.  The client code is backward-compatible.
More information can be found in the @file{README} file!
@item
Fixed some problems when using very long, illegal names.
@end itemize


@node News-3.21.6, News-3.21.5, News-3.21.7, News-3.21.x
@appendixsubsec Changes in release 3.21.6

@itemize @bullet
@item
Fixed more portability issues (incorrect @code{sigwait} and @code{sigset}
defines).
@item
@code{configure} should now be able to detect the last argument to
@code{accept()}.
@end itemize


@node News-3.21.5, News-3.21.4, News-3.21.6, News-3.21.x
@appendixsubsec Changes in release 3.21.5

@itemize @bullet
@item
Should now work with FreeBSD 3.0 if used with
@file{FreeBSD-3.0-libc_r-1.0.diff},
which can be found at @uref{http://www.mysql.com/Downloads/Patches/}.
@item
Added new @code{-O tmp_table_size=#} option to @code{mysqld}.
@item
New function @code{FROM_UNIXTIME(timestamp)} which returns a date string in
@code{'YYYY-MM-DD HH:MM:DD'} format.
@item
New function @code{SEC_TO_TIME(seconds)} which returns a string in
@code{'HH:MM:SS'} format.
@item
New function @code{SUBSTRING_INDEX()}, originally by Zeev Suraski.
@end itemize


@node News-3.21.4, News-3.21.3, News-3.21.5, News-3.21.x
@appendixsubsec Changes in release 3.21.4

@itemize @bullet
@item
Should now configure and compile on OSF/1 4.0 with the DEC compiler.
@item
Configuration and compilation on BSD/OS 3.0 works, but due to some bugs in
BSD/OS 3.0, @code{mysqld} doesn't work on it yet.
@item
Configuration and compilation on FreeBSD 3.0 works, but I couldn't get
@code{pthread_create} to work.
@end itemize


@node News-3.21.3, News-3.21.2, News-3.21.4, News-3.21.x
@appendixsubsec Changes in release 3.21.3

@itemize @bullet
@item
Added reverse check lookup of hostnames to get better security.
@item
Fixed some possible buffer overflows if filenames that are too long are used.
@item
@code{mysqld} doesn't accept hostnames that start with digits followed by a
@code{'.'}, because the hostname may look like an IP number.
@item
Added @code{--skip-networking} option to @code{mysqld}, to allow only socket
connections.  (This will not work with MIT-pthreads!)
@item
Added check of too long table names for alias.
@item
Added check if database name is okay.
@item
Added check if too long table names.
@item
Removed incorrect @code{free()} that killed the server on
@code{CREATE DATABASE} or @code{DROP DATABASE}.
@item
Changed some @code{mysqld} @code{-O} options to better names.
@item
Added @code{-O join_cache_size=#} option to @code{mysqld}.
@item
Added @code{-O max_join_size=#} option to @code{mysqld}, to be able to set a
limit how big queries (in this case big = slow) one should be able to handle
without specifying @code{SET OPTION SQL_BIG_SELECTS=1}.  A # = is about 10
examined records. The default is ``unlimited''.
@item
When comparing a @code{TIME}, @code{DATE}, @code{DATETIME} or @code{TIMESTAMP}
column to a constant, the constant is converted to a time value before
performing the comparison.
This will make it easier to get ODBC (particularly Access97) to work with
the above types. It should also make dates easier to use and the comparisons
should be quicker than before.
@item
Applied patch from Jochen Wiedmann that allows @code{query()} in
@code{mysqlperl} to take a query with @code{\0} in it.
@item
Storing a timestamp with a 2-digit year (@code{YYMMDD}) didn't work.
@item
Fix that timestamp wasn't automatically updated if set in an @code{UPDATE}
clause.
@item
Now the automatic timestamp field is the FIRST timestamp field.
@item
@code{SELECT * INTO OUTFILE}, which didn't correctly if the outfile already
existed.
@item
@code{mysql} now shows the thread ID when starting or doing a reconnect.
@item
Changed the default sort buffer size from 2M to 1M.
@end itemize


@node News-3.21.2, News-3.21.0, News-3.21.3, News-3.21.x
@appendixsubsec Changes in release 3.21.2

@itemize @bullet
@item
The range optimiser is coded, but only 85% tested. It can be enabled with
@code{--new}, but it crashes core a lot yet...
@item
More portable. Should compile on AIX and alpha-digital.
At least the @code{isam} library should be relatively 64-bit clean.
@item
New @code{isamchk} which can detect and fix more problems.
@item
New options for @code{isamlog}.
@item
Using new version of Automake.
@item
Many small portability changes (from the AIX and alpha-digital port)
Better checking of pthread(s) library.
@item
czech error messages by @email{snajdr@@pvt.net}.
@item
Decreased size of some buffers to get fewer problems on systems with little
memory.  Also added more checks to handle ``out of memory'' problems.
@item
@code{mysqladmin}: you can now do @code{mysqladmin kill 5,6,7,8} to kill
multiple threads.
@item
When the maximum connection limit is reached, one extra connection by a user
with the @strong{process_acl} privilege is granted.
@item
Added @code{-O backlog=#} option to @code{mysqld}.
@item
Increased maximum packet size from 512K to 1024K for client.
@item
Almost all of the function code is now tested in the internal test suite.
@item
@code{ALTER TABLE} now returns warnings from field conversions.
@item
Port changed to 3306 (got it reserved from ISI).
@item
Added a fix for Visual FoxBase so that any schema name from a table
specification is automatically removed.
@item
New function @code{ASCII()}.
@item
Removed function @code{BETWEEN(a,b,c)}. Use the standard ANSI
syntax instead: @code{expr BETWEEN expr AND expr}.
@item
MySQL no longer has to use an extra temporary table when sorting
on functions or @code{SUM()} functions.
@item
Fixed bug that you couldn't use @code{tbl_name.field_name} in @code{UPDATE}.
@item
Fixed @code{SELECT DISTINCT} when using 'hidden group'. For example:
@example
mysql> SELECT DISTINCT MOD(some_field,10) FROM test
    ->        GROUP BY some_field;
@end example
Note: @code{some_field} is normally in the @code{SELECT} part. ANSI SQL should
require it.
@end itemize


@node News-3.21.0,  , News-3.21.2, News-3.21.x
@appendixsubsec Changes in release 3.21.0

@itemize @bullet
@item
New reserved words used: @code{INTERVAL}, @code{EXPLAIN}, @code{READ},
@code{WRITE}, @code{BINARY}.
@item
Added ODBC function @code{CHAR(num,...)}.
@item
New operator @code{IN}. This uses a binary search to find a match.
@item
New command @code{LOCK TABLES tbl_name [AS alias] @{READ|WRITE@} ...}
@item
Added @code{--log-update} option to @code{mysqld}, to get a log suitable for
incremental updates.
@item
New command @code{EXPLAIN SELECT ...} to get information about how the
optimiser will do the join.
@item
For easier client code, the client should no longer use
@code{FIELD_TYPE_TINY_BLOB}, @code{FIELD_TYPE_MEDIUM_BLOB},
@code{FIELD_TYPE_LONG_BLOB} or @code{FIELD_TYPE_VAR_STRING} (as
previously returned by @code{mysql_list_fields}).  You should instead only use
@code{FIELD_TYPE_BLOB} or @code{FIELD_TYPE_STRING}.  If you want exact
types, you should use the command @code{SHOW FIELDS}.
@item
Added varbinary syntax: @code{0x######} which can be used as a string
(default) or a number.
@item
@code{FIELD_TYPE_CHAR} is renamed to @code{FIELD_TYPE_TINY}.
@item
Changed all fields to C++ classes.
@item
Removed FORM struct.
@item
Fields with @code{DEFAULT} values no longer need to be @code{NOT NULL}.
@item
New field types:
@table @code
@item ENUM
A string which can take only a couple of defined values. The value is
stored as a 1-3 byte number that is mapped automatically to a string.
This is sorted according to string positions!
@item SET
A string which may have one or many string values separated with ','.
The string is stored as a 1-, 2-, 3-, 4- or 8-byte number where each bit stands
for a specific set member.  This is sorted according to the unsigned value
of the stored packed number.
@end table
@item
Now all function calculation is done with @code{double} or @code{long long}.
This will provide the full 64-bit range with bit functions and fix some
conversions that previously could result in precision losses.  One should
avoid using @code{unsigned long long} columns with full 64-bit range
(numbers bigger than 9223372036854775807) because calculations are done
with @code{signed long long}.
@item
@code{ORDER BY} will now put @code{NULL} field values first. @code{GROUP BY}
will also work with @code{NULL} values.
@item
Full @code{WHERE} with expressions.
@item
New range optimiser that can resolve ranges when some keypart prefix is
constant. Example:
@example
mysql> SELECT * FROM tbl_name
    ->          WHERE key_part_1="customer"
    ->          AND key_part_2>=10 AND key_part_2<=10;
@end example
@end itemize


@node News-3.20.x, News-3.19.x, News-3.21.x, News
@appendixsec Changes in release 3.20.x

@cindex changes, version 3.20

Version 3.20 is quite old now, and should be avoided if possible.  This
information is kept here for historical purposes only.

Changes from 3.20.18 to 3.20.32b are not documented here because the
3.21 release branched here. And the relevant changes are also
documented as changes to the 3.21 version.

@menu
* News-3.20.18::                Changes in release 3.20.18
* News-3.20.17::                Changes in release 3.20.17
* News-3.20.16::                Changes in release 3.20.16
* News-3.20.15::                Changes in release 3.20.15
* News-3.20.14::                Changes in release 3.20.14
* News-3.20.13::                Changes in release 3.20.13
* News-3.20.11::                Changes in release 3.20.11
* News-3.20.10::                Changes in release 3.20.10
* News-3.20.9::                 Changes in release 3.20.9
* News-3.20.8::                 Changes in release 3.20.8
* News-3.20.7::                 Changes in release 3.20.7
* News-3.20.6::                 Changes in release 3.20.6
* News-3.20.3::                 Changes in release 3.20.3
* News-3.20.0::                 Changes in releases 3.20.0
@end menu


@node News-3.20.18, News-3.20.17, News-3.20.x, News-3.20.x
@appendixsubsec Changes in release 3.20.18

@itemize @bullet
@item
Added @code{-p#} (remove @code{#} directories from path) to @code{isamlog}.
All files are written with a relative path from the database directory
Now @code{mysqld} shouldn't crash on shutdown when using the
@code{--log-isam} option.
@item
New @code{mysqlperl} version. It is now compatible with @code{msqlperl-0.63}.
@item
New @code{DBD} module available at @uref{http://www.mysql.com/Downloads/Contrib/}
site.
@item
Added group function @code{STD()} (standard deviation).
@item
The @code{mysqld} server is now compiled by default without debugging
information.  This will make the daemon smaller and faster.
@item
Now one usually only has to specify the @code{--basedir} option to
@code{mysqld}. All other paths are relative in a normal installation.
@item
@code{BLOB} columns sometimes contained garbage when used with a @code{SELECT}
on more than one table and @code{ORDER BY}.
@item
Fixed that calculations that are not in @code{GROUP BY} work as expected
(ANSI SQL extension).
Example:
@example
mysql> SELECT id,id+1 FROM table GROUP BY id;
@end example
@item
The test of using @code{MYSQL_PWD} was reversed. Now @code{MYSQL_PWD} is
enabled as default in the default release.
@item
Fixed conversion bug which caused @code{mysqld} to core dump with
Arithmetic error on SPARC-386.
@item
Added @code{--unbuffered} option to @code{mysql}, for new @code{mysqlaccess}.
@item
When using overlapping (unnecessary) keys and join over many tables,
the optimiser could get confused and return 0 records.
@end itemize


@node News-3.20.17, News-3.20.16, News-3.20.18, News-3.20.x
@appendixsubsec Changes in release 3.20.17

@itemize @bullet
@item
You can now use @code{BLOB} columns and the functions @code{IS NULL} and
@code{IS NOT NULL} in the @code{WHERE} clause.
@item
All communication packets and row buffers are now allocated dynamically
on demand. The default value of @code{max_allowed_packet} is now 64K for
the server and 512K for the client. This is mainly used to catch
incorrect packets that could trash all memory. The server limit may be
changed when it is started.
@item
Changed stack usage to use less memory.
@item
Changed @code{safe_mysqld} to check for running daemon.
@item
The @code{ELT()} function is renamed to @code{FIELD()}.  The new
@code{ELT()} function returns a value based on an index: @code{FIELD()}
is the inverse of @code{ELT()} Example: @code{ELT(2,"A","B","C")} returns
@code{"B"}. @code{FIELD("B","A","B","C")} returns @code{2}.
@item
@code{COUNT(field)}, where @code{field} could have a @code{NULL} value, now
works.
@item
A couple of bugs fixed in @code{SELECT ... GROUP BY}.
@item
Fixed memory overrun bug in @code{WHERE} with many unoptimisable brace levels.
@item
Fixed some small bugs in the grant code.
@item
If hostname isn't found by @code{get_hostname}, only the IP is checked.
Previously, you got @code{Access denied}.
@item
Inserts of timestamps with values didn't always work.
@item
@code{INSERT INTO ... SELECT ... WHERE} could give the error
@code{Duplicated field}.
@item
Added some tests to @code{safe_mysqld} to make it ``safer''.
@item
@code{LIKE} was case-sensitive in some places and case-insensitive in others.
Now @code{LIKE} is always case-insensitive.
@item
@file{mysql.cc}: Allow @code{'#'} anywhere on the line.
@item
New command @code{SET OPTION SQL_SELECT_LIMIT=#}. See the FAQ for more details.
@item
New version of the @code{mysqlaccess} script.
@item
Change @code{FROM_DAYS()} and @code{WEEKDAY()} to also take a full
@code{TIMESTAMP} or @code{DATETIME} as argument. Before they only took a
number of type @code{YYYYMMDD} or @code{YYMMDD}.
@item
Added new function @code{UNIX_TIMESTAMP(timestamp_column)}.
@end itemize


@node News-3.20.16, News-3.20.15, News-3.20.17, News-3.20.x
@appendixsubsec Changes in release 3.20.16

@itemize @bullet
@item
More changes in MIT-pthreads to get them safer. Fixed also some link
bugs at least in SunOS.
@item
Changed @code{mysqld} to work around a bug in MIT-pthreads. This makes multiple
small @code{SELECT} operations 20 times faster. Now @code{lock_test.pl} should
work.
@item
Added @code{mysql_FetchHash(handle)} to @code{mysqlperl}.
@item
The @code{mysqlbug} script is now distributed built to allow for reporting
bugs that appear during the build with it.
@item
Changed @file{libmysql.c} to prefer @code{getpwuid()} instead of
@code{cuserid()}.
@item
Fixed bug in @code{SELECT} optimiser when using many tables with the same
column used as key to different tables.
@item
Added new @code{latin2} and Russian @code{KOI8} character tables.
@item
Added support for a dummy @code{GRANT} command to satisfy Powerbuilder.
@end itemize


@node News-3.20.15, News-3.20.14, News-3.20.16, News-3.20.x
@appendixsubsec Changes in release 3.20.15

@itemize @bullet
@item
Fixed fatal bug @code{packets out of order} when using MIT-pthreads.
@item
Removed possible loop when a thread waits for command from client
and @code{fcntl()} fails. Thanks to Mike Bretz for finding this bug.
@item
Changed alarm loop in @file{mysqld.cc} because shutdown didn't always
succeed in Linux.
@item
Removed use of @code{termbits} from @file{mysql.cc}.  This conflicted with
@code{glibc} 2.0.
@item
Fixed some syntax errors for at least BSD and Linux.
@item
Fixed bug when doing a @code{SELECT} as superuser without a database.
@item
Fixed bug when doing @code{SELECT} with group calculation to outfile.
@end itemize


@node News-3.20.14, News-3.20.13, News-3.20.15, News-3.20.x
@appendixsubsec Changes in release 3.20.14

@itemize @bullet
@item
If one gives @code{-p} or @code{--password} option to @code{mysql} without
an argument, the user is solicited for the password from the tty.
@item
Added default password from @code{MYSQL_PWD} (by Elmar Haneke).
@item
Added command @code{kill} to @code{mysqladmin} to kill a specific
MySQL thread.
@item
Sometimes when doing a reconnect on a down connection this succeeded
first on second try.
@item
Fixed adding an @code{AUTO_INCREMENT} key with @code{ALTER_TABLE}.
@item
@code{AVG()} gave too small value on some @code{SELECT}s with
@code{GROUP BY} and @code{ORDER BY}.
@item
Added new @code{DATETIME} type (by Giovanni Maruzzelli
@email{maruzz@@matrice.it}).
@item
Fixed that defining @code{DONT_USE_DEFAULT_FIELDS} works.
@item
Changed to use a thread to handle alarms instead of signals on Solaris to
avoid race conditions.
@item
Fixed default length of signed numbers. (George Harvey
@email{georgeh@@pinacl.co.uk}.)
@item
Allow anything for @code{CREATE INDEX}.
@item
Add prezeros when packing numbers to @code{DATE}, @code{TIME} and
@code{TIMESTAMP}.
@item
Fixed a bug in @code{OR} of multiple tables (gave empty set).
@item
Added many patches to MIT-pthreads. This fixes at least one lookup bug.
@end itemize


@node News-3.20.13, News-3.20.11, News-3.20.14, News-3.20.x
@appendixsubsec Changes in release 3.20.13

@itemize @bullet
@item
Added ANSI SQL94 @code{DATE} and @code{TIME} types.
@item
Fixed bug in @code{SELECT} with @code{AND}-@code{OR} levels.
@item
Added support for Slovenian characters. The @file{Contrib} directory contains
source and instructions for adding other character sets.
@item
Fixed bug with @code{LIMIT} and @code{ORDER BY}.
@item
Allow @code{ORDER BY} and @code{GROUP BY} on items that aren't in the
@code{SELECT} list.
(Thanks to Wim Bonis @email{bonis@@kiss.de}, for pointing this out.)
@item
Allow setting of timestamp values in @code{INSERT}.
@item
Fixed bug with @code{SELECT ... WHERE ... = NULL}.
@item
Added changes for @code{glibc} 2.0. To get @code{glibc} to work, you should
add the @file{gibc-2.0-sigwait-patch} before compiling @code{glibc}.
@item
Fixed bug in @code{ALTER TABLE} when changing a @code{NOT NULL} field to
allow @code{NULL} values.
@item
Added some ANSI92 synonyms as field types to @code{CREATE TABLE}.
@code{CREATE TABLE} now allows @code{FLOAT(4)} and @code{FLOAT(8)} to mean
@code{FLOAT} and @code{DOUBLE}.
@item
New utility program @code{mysqlaccess} by @email{Yves.Carlier@@rug.ac.be}.
This program shows the access rights for a specific user and the grant
rows that determine this grant.
@item
Added @code{WHERE const op field} (by @email{bonis@@kiss.de}).
@end itemize


@node News-3.20.11, News-3.20.10, News-3.20.13, News-3.20.x
@appendixsubsec Changes in release 3.20.11

@itemize @bullet
@item
When using @code{SELECT ... INTO OUTFILE}, all temporary tables are ISAM
instead of HEAP to allow big dumps.
@item
Changed date functions to be string functions. This fixed some ``funny''
side effects when sorting on dates.
@item
Extended @code{ALTER TABLE} according to SQL92.
@item
Some minor compatibility changes.
@item
Added @code{--port} and @code{--socket} options to all utility programs and
@code{mysqld}.
@item
Fixed MIT-pthreads @code{readdir_r()}. Now @code{mysqladmin create database}
and @code{mysqladmin drop database} should work.
@item
Changed MIT-pthreads to use our @code{tempnam()}. This should fix the ``sort
aborted'' bug.
@item
Added sync of records count in @code{sql_update}. This fixed slow updates
on first connection. (Thanks to Vaclav Bittner for the test.)
@end itemize


@node News-3.20.10, News-3.20.9, News-3.20.11, News-3.20.x
@appendixsubsec Changes in release 3.20.10

@itemize @bullet
@item
New insert type: @code{INSERT INTO ... SELECT ...}
@item
@code{MEDIUMBLOB} fixed.
@item
Fixed bug in @code{ALTER TABLE} and @code{BLOB}s.
@item
@code{SELECT ... INTO OUTFILE} now creates the file in the current
database directory.
@item
@code{DROP TABLE} now can take a list of tables.
@item
Oracle synonym @code{DESCRIBE} (@code{DESC}).
@item
Changes to @code{make_binary_distribution}.
@item
Added some comments to installation instructions about @code{configure}'s
C++ link test.
@item
Added @code{--without-perl} option to @code{configure}.
@item
Lots of small portability changes.
@end itemize


@node News-3.20.9, News-3.20.8, News-3.20.10, News-3.20.x
@appendixsubsec Changes in release 3.20.9

@itemize @bullet
@item
@code{ALTER TABLE} didn't copy null bit. As a result, fields that were allowed
to have @code{NULL} values were always @code{NULL}.
@item
@code{CREATE} didn't take numbers as @code{DEFAULT}.
@item
Some compatibility changes for SunOS.
@item
Removed @file{config.cache} from old distribution.
@end itemize


@node News-3.20.8, News-3.20.7, News-3.20.9, News-3.20.x
@appendixsubsec Changes in release 3.20.8

@itemize @bullet
@item
Fixed bug with @code{ALTER TABLE} and multi-part keys.
@end itemize


@node News-3.20.7, News-3.20.6, News-3.20.8, News-3.20.x
@appendixsubsec Changes in release 3.20.7

@itemize @bullet
@item
New commands: @code{ALTER TABLE}, @code{SELECT ... INTO OUTFILE} and
@code{LOAD DATA INFILE}.
@item
New function: @code{NOW()}.
@item
Added new field @code{File_priv} to @code{mysql/user} table.
@item
New script @code{add_file_priv} which adds the new field @code{File_priv}
to the @code{user} table. This script must be executed if you want to
use the new @code{SELECT ... INTO} and @code{LOAD DATA INFILE ...} commands
with a version of MySQL earlier than 3.20.7.
@item
Fixed bug in locking code, which made @code{lock_test.pl} test fail.
@item
New files @file{NEW} and @file{BUGS}.
@item
Changed @file{select_test.c} and @file{insert_test.c} to include
@file{config.h}.
@item
Added @code{status} command to @code{mysqladmin} for short logging.
@item
Increased maximum number of keys to 16 and maximum number of key parts to 15.
@item
Use of sub keys. A key may now be a prefix of a string field.
@item
Added @code{-k} option to @code{mysqlshow}, to get key information for a table.
@item
Added long options to @code{mysqldump}.
@end itemize


@node News-3.20.6, News-3.20.3, News-3.20.7, News-3.20.x
@appendixsubsec Changes in release 3.20.6

@itemize @bullet
@item
Portable to more systems because of MIT-pthreads, which will
be used automatically if @code{configure} cannot find a @code{-lpthreads}
library.
@item
Added GNU-style long options to almost all programs. Test with
@code{@kbd{program} --help}.
@item
Some shared library support for Linux.
@item
The FAQ is now in @file{.texi} format and is available in @file{.html},
@file{.txt} and @file{.ps} formats.
@item
Added new SQL function @code{RAND([init])}.
@item
Changed @code{sql_lex} to handle @code{\0} unquoted, but the client can't send
the query through the C API, because it takes a str pointer.
You must use @code{mysql_real_query()} to send the query.
@item
Added API function @code{mysql_get_client_info()}.
@item
@code{mysqld} now uses the @code{N_MAX_KEY_LENGTH} from @file{nisam.h} as
the maximum allowable key length.
@item
The following now works:
@example
mysql> SELECT filter_nr,filter_nr FROM filter ORDER BY filter_nr;
@end example
Previously, this resulted in the error:
@code{Column: 'filter_nr' in order clause is ambiguous}.
@item
@code{mysql} now outputs @code{'\0'}, @code{'\t'}, @code{'\n'} and @code{'\\'}
when encountering ASCII 0, tab, newline or @code{'\'} while writing
tab-separated output.
This is to allow printing of binary data in a portable format.
To get the old behavior, use @code{-r} (or @code{--raw}).
@item
Added german error messages (60 of 80 error messages translated).
@item
Added new API function @code{mysql_fetch_lengths(MYSQL_RES *)}, which
returns an array of column lengths (of type @code{uint}).
@item
Fixed bug with @code{IS NULL} in @code{WHERE} clause.
@item
Changed the optimiser a little to get better results when searching on a key
part.
@item
Added @code{SELECT} option @code{STRAIGHT_JOIN} to tell the optimiser that
it should join tables in the given order.
@item
Added support for comments starting with @code{'--'} in @file{mysql.cc}
(Postgres syntax).
@item
You can have @code{SELECT} expressions and table columns in a @code{SELECT}
which are not used in the group part. This makes it efficient to implement
lookups.  The column that is used should be a constant for each group because
the value is calculated only once for the first row that is found for a group.
@example
mysql> SELECT id,lookup.text,SUM(*) FROM test,lookup
    ->        WHERE test.id=lookup.id GROUP BY id;
@end example
@item
Fixed bug in @code{SUM(function)} (could cause a core dump).
@item
Changed @code{AUTO_INCREMENT} placement in the SQL query:
@example
INSERT INTO table (auto_field) VALUES (0);
@end example
inserted 0, but it should insert an @code{AUTO_INCREMENT} value.
@item
@file{mysqlshow.c}: Added number of records in table. Had to change the
client code a little to fix this.
@item
@code{mysql} now allows doubled @code{''} or @code{""} within strings for
embedded @code{'} or @code{"}.
@item
New math functions:
@code{EXP()}, @code{LOG()}, @code{SQRT()}, @code{ROUND()}, @code{CEILING()}.
@end itemize


@node News-3.20.3, News-3.20.0, News-3.20.6, News-3.20.x
@appendixsubsec Changes in release 3.20.3

@itemize @bullet
@item
The @code{configure} source now compiles a thread-free client library
@code{-lmysqlclient}. This is the only library that needs to be linked
with client applications.  When using the binary releases, you must
link with @code{-lmysql -lmysys -ldbug -lmystrings} as before.
@item
New @code{readline} library from @code{bash-2.0}.
@item
LOTS of small changes to @code{configure} and makefiles (and related source).
@item
It should now be possible to compile in another directory using
@code{VPATH}. Tested with GNU Make 3.75.
@item
@code{safe_mysqld} and @code{mysql.server} changed to be more compatible
between the source and the binary releases.
@item
@code{LIMIT} now takes one or two numeric arguments.
If one argument is given, it indicates the maximum number of rows in
a result. If two arguments are given, the first argument indicates the offset
of the first row to return, the second is the maximum number of rows.
With this it's easy to do a poor man's next page/previous page WWW
application.
@item
Changed name of SQL function @code{FIELDS()} to @code{ELT()}.
Changed SQL function @code{INTERVALL()} to @code{INTERVAL()}.
@item
Made @code{SHOW COLUMNS} a synonym for @code{SHOW FIELDS}.
Added compatibility syntax @code{FRIEND KEY} to @code{CREATE TABLE}. In
MySQL, this creates a non-unique key on the given columns.
@item
Added @code{CREATE INDEX} and @code{DROP INDEX} as compatibility functions.
In MySQL, @code{CREATE INDEX} only checks if the index exists and
issues an error if it doesn't exist. @code{DROP INDEX} always succeeds.
@item
@file{mysqladmin.c}: added client version to version information.
@item
Fixed core dump bug in @code{sql_acl} (core on new connection).
@item
Removed @code{host}, @code{user} and @code{db} tables from database @code{test}
in the distribution.
@item
@code{FIELD_TYPE_CHAR} can now be signed (-128 to 127) or unsigned (0 to 255)
Previously, it was always unsigned.
@item
Bug fixes in @code{CONCAT()} and @code{WEEKDAY()}.
@item
Changed a lot of source to get @code{mysqld} to be compiled with SunPro
compiler.
@item
SQL functions must now have a @code{'('} immediately after the function name
(no intervening space).
For example, @code{'USER('} is regarded as beginning a function call, and
@code{'USER ('} is regarded as an identifier @code{USER} followed by a
@code{'('}, not as a function call.
@end itemize


@node News-3.20.0,  , News-3.20.3, News-3.20.x
@appendixsubsec Changes in release 3.20.0

@itemize @bullet
@item
The source distribution is done with @code{configure} and Automake.
It will make porting much easier.  The @code{readline} library is included
in the distribution.
@item
Separate client compilation:
the client code should be very easy to compile on systems which
don't have threads.
@item
The old Perl interface code is automatically compiled and installed.
Automatic compiling of @code{DBD} will follow when the new @code{DBD} code
is ported.
@item
Dynamic language support: @code{mysqld} can now be started with Swedish
or English (default) error messages.
@item
New functions: @code{INSERT()}, @code{RTRIM()}, @code{LTRIM()} and
@code{FORMAT()}.
@item
@code{mysqldump} now works correctly for all field types (even
@code{AUTO_INCREMENT}). The format for @code{SHOW FIELDS FROM tbl_name}
is changed so the @code{Type} column contains information suitable for
@code{CREATE TABLE}.  In previous releases, some @code{CREATE TABLE}
information had to be patched when re-creating tables.
@item
Some parser bugs from 3.19.5 (@code{BLOB} and @code{TIMESTAMP}) are corrected.
@code{TIMESTAMP} now returns different date information depending on its
create length.
@item
Changed parser to allow a database, table or field name to
start with a number or @code{'_'}.
@item
All old C code from Unireg changed to C++ and cleaned up. This makes
the daemon a little smaller and easier to understand.
@item
A lot of small bug fixes done.
@item
New @file{INSTALL} files (not final version) and some information regarding porting.
@end itemize


@node News-3.19.x,  , News-3.20.x, News
@appendixsec Changes in release 3.19.x

@cindex changes, version 3.19

Version 3.19 is quite old now, and should be avoided if possible.  This
information is kept here for historical purposes only.

@menu
* News-3.19.5::                 Changes in release 3.19.5
* News-3.19.4::                 Changes in release 3.19.4
* News-3.19.3::                 Changes in release 3.19.3
@end menu


@node News-3.19.5, News-3.19.4, News-3.19.x, News-3.19.x
@appendixsubsec Changes in release 3.19.5

@itemize @bullet
@item
Some new functions, some more optimisation on joins.
@item
Should now compile clean on Linux (2.0.x).
@item
Added functions @code{DATABASE()}, @code{USER()}, @code{POW()},
@code{LOG10()} (needed for ODBC).
@item
In a @code{WHERE} with an @code{ORDER BY} on fields from only one table,
the table is now preferred as first table in a multi-join.
@item
@code{HAVING} and @code{IS NULL} or @code{IS NOT NULL} now works.
@item
A group on one column and a sort on a group function (@code{SUM()},
@code{AVG()}...) didn't work together. Fixed.
@item
@code{mysqldump}: Didn't send password to server.
@end itemize


@node News-3.19.4, News-3.19.3, News-3.19.5, News-3.19.x
@appendixsubsec Changes in release 3.19.4

@itemize @bullet
@item
Fixed horrible locking bug when inserting in one thread and reading
in another thread.
@item
Fixed one-off decimal bug.  1.00 was output as 1.0.
@item
Added attribute @code{'Locked'} to process list as info if a query is
locked by another query.
@item
Fixed full magic timestamp. Timestamp length may now be 14, 12, 10, 8, 6, 4
or 2 bytes.
@item
Sort on some numeric functions could sort incorrectly on last number.
@item
@code{IF(arg,syntax_error,syntax_error)} crashed.
@item
Added functions @code{CEILING()}, @code{ROUND()}, @code{EXP()}, @code{LOG()} and @code{SQRT()}.
@item
Enhanced @code{BETWEEN} to handle strings.
@end itemize


@node News-3.19.3,  , News-3.19.4, News-3.19.x
@appendixsubsec Changes in release 3.19.3

@itemize @bullet
@item
Fixed @code{SELECT} with grouping on @code{BLOB} columns not to return
incorrect @code{BLOB} info. Grouping, sorting and distinct on @code{BLOB}
columns will not yet work as
expected (probably it will group/sort by the first 7 characters in the
@code{BLOB}). Grouping on formulas with a fixed string size (use @code{MID()}
on a @code{BLOB}) should work.
@item
When doing a full join (no direct keys) on multiple tables with @code{BLOB}
fields, the @code{BLOB} was garbage on output.
@item
Fixed @code{DISTINCT} with calculated columns.
@end itemize





@node Porting, Environment variables, News, Top
@appendix Porting to Other Systems

@cindex porting, to other systems

This appendix will help you port MySQL to other operationg systems.
Do check the list of currently supported operating systems first.
@xref{Which OS}.
If you have created a new port of MySQL, please let us know so that
we can list it here and on our web site (@uref{http://www.mysql.com/}),
recommending it to other users.

Note: If you create a new port of MySQL, you are free to copy and
distribute it under the GPL license, but it does not make you a
copyright holder of MySQL.

A working Posix thread library is needed for the server. On Solaris 2.5
we use Sun PThreads (the native thread support in 2.4 and earlier
versions are not good enough) and on Linux we use LinuxThreads by Xavier
Leroy, @email{Xavier.Leroy@@inria.fr}.

The hard part of porting to a new Unix variant without good native
thread support is probably to port MIT-pthreads. See
@file{mit-pthreads/README} and Programming POSIX Threads
(@uref{http://www.humanfactor.com/pthreads/}).

The MySQL distribution includes a patched version of
Provenzano's Pthreads from MIT (see the MIT Pthreads web page at
@uref{http://www.mit.edu:8001/people/proven/pthreads.html}).
This can be used for some operating systems that do not
have POSIX threads.

It is also possible to use another user level thread package named
FSU Pthreads (see
@uref{http://www.informatik.hu-berlin.de/~mueller/pthreads.html, FSU Pthreads
home page}). This implementation is being used for the SCO port.

See the @file{thr_lock.c} and @file{thr_alarm.c} programs in the @file{mysys}
directory for some tests/examples of these problems.

Both the server and the client need a working C++ compiler (we use @code{gcc}
and have tried SPARCworks). Another compiler that is known to work is the
Irix @code{cc}.

To compile only the client use @code{./configure --without-server}.

There is currently no support for only compiling the server, nor is it
likly to be added unless someone has a good reason for it.

If you want/need to change any @file{Makefile} or the configure script you must
get Automake and Autoconf. We have used the @code{automake-1.2} and
@code{autoconf-2.12} distributions.

All steps needed to remake everything from the most basic files.

@example
/bin/rm */.deps/*.P
/bin/rm -f config.cache
aclocal
autoheader
aclocal
automake
autoconf
./configure --with-debug=full --prefix='your installation directory'

# The makefiles generated above need GNU make 3.75 or newer.
# (called gmake below)
gmake clean all install init-db
@end example

If you run into problems with a new port, you may have to do some debugging
of MySQL!
@xref{Debugging server}.

@strong{Note}: before you start debugging @code{mysqld}, first get the test
programs @code{mysys/thr_alarm} and @code{mysys/thr_lock} to work.  This
will ensure that your thread installation has even a remote chance to work!

@menu
* Debugging server::            Debugging a MySQL server
* Debugging client::            Debugging a MySQL client
* The DBUG package::            The DBUG Package
* Locking methods::             Locking ethods
* RTS-threads::                 Comments about RTS threads
* Thread packages::             Differences between different thread packages
@end menu


@node Debugging server, Debugging client, Porting, Porting
@appendixsec Debugging a MySQL server

@cindex server, debugging
@cindex debugging, server
@cindex crash

If you are using some functionality that is very new in MySQL,
you can try to run @code{mysqld} with the @code{--skip-new} (which will disable all
new, potentially unsafe functionality) or with @code{--safe-mode} which
disables a lot of optimisation that may cause problems.
@xref{Crashing}.

If @code{mysqld} doesn't want to start, you should check that you don't have
any @file{my.cnf} files that interfere with your setup!
You can check your @file{my.cnf} arguments with @code{mysqld --print-defaults}
and avoid using them by starting with @code{mysqld --no-defaults ...}.

If @code{mysqld} starts to eat up CPU or memory or if it ``hangs'', you
can use @code{mysqladmin processlist status} to find out if someone is
executing a query that takes a long time.  It may be a good idea to
run @code{mysqladmin -i10 processlist status} in some window if you are
experiencing performance problems or problems when new clients can't connect.

The command @code{mysqladmin debug} will dump some information about
locks in use, used memory and query usage to the mysql log file. This
may help solve some problems.  This command also provides some useful
information even if you haven't compiled MySQL for debugging!

If the problem is that some tables are getting slower and slower you
should try to optimise the table with @code{OPTIMIZE TABLE} or
@code{myisamchk}.  @xref{MySQL Database Administration}. You should also
check the slow queries with @code{EXPLAIN}.

You should also read the OS-specific section in this manual for
problems that may be unique to your environment.
@xref{Operating System Specific Notes}.

@menu
* Compiling for debugging::     Compiling MYSQL for debugging
* Making trace files::          Creating trace files
* Using gdb on mysqld::         Debugging mysqld under gdb
* Using stack trace::           Using a stack trace
* Using log files::             Using log files to find cause of errors in mysqld
* Reproduceable test case::     Making a test case when you experience table corruption
@end menu


@node Compiling for debugging, Making trace files, Debugging server, Debugging server
@appendixsubsec Compiling MYSQL for Debugging

If you have some very specific problem, you can always try to debug
MySQL.  To do this you must configure MySQL with the
@code{--with-debug} or the @code{--with-debug=full} option.  You can check
whether MySQL was compiled with debugging by doing:
@code{mysqld --help}.  If the @code{--debug} flag is listed with the
options then you have debugging enabled.  @code{mysqladmin ver} also
lists the @code{mysqld} version as @code{mysql ... --debug} in this case.

If you are using gcc or egcs, the recommended configure line is:

@example
CC=gcc CFLAGS="-O2" CXX=gcc CXXFLAGS="-O2 -felide-constructors \
   -fno-exceptions -fno-rtti" ./configure --prefix=/usr/local/mysql \
   --with-debug --with-extra-charsets=complex
@end example

This will avoid problems with the @code{libstdc++} library and with C++
exceptions (many compilers have problems with C++ exceptions in threaded
code) and compile a MySQL version with support for all character sets.

If you suspect a memory overrun error, you can configure MySQL
with @code{--with-debug=full}, which will install a memory allocation
(@code{SAFEMALLOC}) checker. Running with @code{SAFEMALLOC} is however
quite slow, so if you get performance problems you should start
@code{mysqld} with the @code{--skip-safemalloc} option. This will
disable the memory overrun checks for each call to @code{malloc} and
@code{free}.

If @code{mysqld} stops crashing when you compile it with
@code{--with-debug}, you have probably found a compiler bug or a timing
bug within MySQL.  In this case you can try to add @code{-g} to
the @code{CFLAGS} and @code{CXXFLAGS} variables above and not use
@code{--with-debug}. If @code{mysqld} now dies, you can at least attach
to it with @code{gdb} or use @code{gdb} on the core file to find out
what happened.

When you configure MySQL for debugging you automatically enable a
lot of extra safety check functions that monitor the health of @code{mysqld}.
If they find something ``unexpected,'' an entry will be written to
@code{stderr}, which @code{safe_mysqld} directs to the error log!  This also
means that if you are having some unexpected problems with MySQL and
are using a source distribution, the first thing you should do is to
configure MySQL for debugging!  (The second thing, of course, is to
send mail to @email{mysql@@lists.mysql.com} and ask for help.  Please use the
@code{mysqlbug} script for all bug reports or questions regarding the
MySQL version you are using!

In the Windows MySQL distribution, @code{mysqld.exe} is by
default compiled with support for trace files.


@node Making trace files, Using gdb on mysqld, Compiling for debugging, Debugging server
@appendixsubsec Creating Trace Files

If the @code{mysqld} server doesn't start or if you can cause the
@code{mysqld} server to crash quickly, you can try to create a trace
file to find the problem.

To do this you have to have a @code{mysqld} that is compiled for debugging.
You can check this by executing @code{mysqld -V}.  If the version number
ends with @code{-debug}, it's compiled with support for trace files.

Start the @code{mysqld} server with a trace log in @file{/tmp/mysqld.trace}
(or @file{C:\mysqld.trace} on Windows):

@code{mysqld --debug}

On Windows you should also use the @code{--standalone} flag to not start
@code{mysqld} as a service:

In a DOS window do:

@example
mysqld --debug --standalone
@end example

After this you can use the @code{mysql.exe} command-line tool in a
second DOS window to reproduce the problem. You can take down the above
@code{mysqld} server with @code{mysqladmin shutdown}.

Note that the trace file will get @strong{very big}!
If you want to have a smaller trace file, you can use something like:

@code{mysqld --debug=d,info,error,query,general,where:O,/tmp/mysqld.trace}

which only prints information with the most interesting tags in
@file{/tmp/mysqld.trace}.

If you make a bug report about this, please only send the lines from the
trace file to the appropriate mailing list where something seems to go
wrong! If you can't locate the wrong place, you can ftp the trace file,
together with a full bug report, to
@uref{ftp://support.mysql.com/pub/mysql/secret/} so that a MySQL
developer can take a look a this.

The trace file is made with the @strong{DBUG} package by Fred Fish.
@xref{The DBUG package}.


@node Using gdb on mysqld, Using stack trace, Making trace files, Debugging server
@appendixsubsec Debugging mysqld under gdb

@cindex gdb, using

On most systems you can also start @code{mysqld} from @code{gdb} to get
more information if @code{mysqld} crashes.

With some older @code{gdb} versions on Linux you must use @code{run
--one-thread} if you want to be able to debug @code{mysqld} threads.  In
this case you can only have one thread active at a time.  We recommend you
to upgrade to gdb 5.1 ASAP as thread debugging works much better with this
version!

When running @code{mysqld} under gdb, you should disable the stack trace
with @code{--skip-stack-trace} to be able to catch segfaults within gdb.

It's very hard to debug MySQL under @code{gdb} if you do a lot of
new connections the whole time as @code{gdb} doesn't free the memory for
old threads.  You can avoid this problem by starting @code{mysqld} with
@code{-O thread_cache_size= 'max_connections +1'}.  In most cases just
using @code{-O thread_cache_size=5'} will help a lot!

If you want to get a core dump on Linux if @code{mysqld} dies with a
SIGSEGV signal, you can start @code{mysqld} with the @code{--core-file} option.
This core file can be used to make a backtrace that may help you
find out why @code{mysqld} died:

@example
shell> gdb mysqld core
gdb>   backtrace full
gdb>   exit
@end example

@xref{Crashing}.

If you are using gdb 4.17.x or above on Linux, you should install a
@file{.gdb} file, with the following information, in your current
directory:

@example
set print sevenbit off
handle SIGUSR1 nostop noprint
handle SIGUSR2 nostop noprint
handle SIGWAITING nostop noprint
handle SIGLWP nostop noprint
handle SIGPIPE nostop
handle SIGALRM nostop
handle SIGHUP nostop
handle SIGTERM nostop noprint
@end example

If you have problems debugging threads with gdb, you should download
gdb 5.x and try this instead. The new gdb version has very improved
thread handling!

Here is an example how to debug mysqld:

@example
shell> gdb /usr/local/libexec/mysqld
gdb> run
...
backtrace full # Do this when mysqld crashes
@end example

Include the above output in a mail generated with @code{mysqlbug} and
mail this to @code{mysql@@lists.mysql.com}.

If @code{mysqld} hangs you can try to use some system tools like
@code{strace} or @code{/usr/proc/bin/pstack} to examine where
@code{mysqld} has hung.

@example
strace /tmp/log libexec/mysqld
@end example

@findex DBI->trace
@findex trace DBI method
@tindex DBI_TRACE environment variable
@tindex Environment variable, DBI_TRACE
If you are using the Perl @code{DBI} interface, you can turn on
debugging information by using the @code{trace} method or by
setting the @code{DBI_TRACE} environment variable.
@xref{Perl DBI Class, , Perl @code{DBI} Class}.


@node Using stack trace, Using log files, Using gdb on mysqld, Debugging server
@appendixsubsec Using a Stack Trace

On some operating systems, the error log will contain a stack trace if
@code{mysqld} dies unexpectedly. You can use this to find out where (and
maybe why) @code{mysqld} died. @xref{Error log}.  To get a stack trace,
you must not compile @code{mysqld} with the @code{-fomit-frame-pointer}
option to gcc. @xref{Compiling for debugging}.

If the error file contains something like the following:

@example
mysqld got signal 11;
The manual section 'Debugging a MySQL server' tells you how to use a
stack trace and/or the core file to produce a readable backtrace that may
help in finding out why mysqld died
Attemping backtrace. You can use the following information to find out
where mysqld died.  If you see no messages after this, something went
terribly wrong
stack range sanity check, ok, backtrace follows
0x40077552
0x81281a0
0x8128f47
0x8127be0
0x8127995
0x8104947
0x80ff28f
0x810131b
0x80ee4bc
0x80c3c91
0x80c6b43
0x80c1fd9
0x80c1686
@end example

you can find where @code{mysqld} died by doing the following:

@enumerate
@item
Copy the above numbers to a file, for example @file{mysqld.stack}.

@item
Make a symbol file for the @code{mysqld} server:
@example
nm -n libexec/mysqld > /tmp/mysqld.sym
@end example

Note that many MySQL binary distributions comes with the above
file, named @code{mysqld.sym.gz}.  In this case you must unpack this by
doing:
@example
gunzip < bin/mysqld.sym.gz > /tmp/mysqld.sym
@end example

@item
Execute @code{resolve_stack_dump -s /tmp/mysqld.sym -n mysqld.stack}.

This will print out where @code{mysqld} died. If this doesn't help you
find out why @code{mysqld} died, you should make a bug report and include
the output from the above commend with the bug report.

Note however that in most cases it will not help us to just have a stack
trace to find the reason for the problem.  To be able to locate the bug
or provide a workaround, we would in most cases need to know the query
that killed @code{mysqld} and preferable a test case so that we can
repeat the problem!  @xref{Bug reports}.
@end enumerate


@node Using log files, Reproduceable test case, Using stack trace, Debugging server
@appendixsubsec Using Log Files to Find Cause of Errors in mysqld

Note that before starting @code{mysqld} with @code{--log} you should
check all your tables with @code{myisamchk}.
@xref{MySQL Database Administration}.

If @code{mysqld} dies or hangs, you should start @code{mysqld} with
@code{--log}.  When @code{mysqld} dies again, you can examine the end of
the log file for the query that killed @code{mysqld}.

If you are using @code{--log} without a file name, the log is stored in
the database directory as 'hostname'.log In most cases it's the last
query in the log file that killed @code{mysqld}, but if possible you
should verify this by restarting @code{mysqld} and executing the found
query from the @code{mysql} command-line tools. If this works, you
should also test all complicated queries that didn't complete.

You can also try the command @code{EXPLAIN} on all @code{SELECT}
statements that takes a long time to ensure that @code{mysqld} is using
indexes properly. @xref{EXPLAIN, , @code{EXPLAIN}}.

You can find the queries that take a long time to execute by starting
@code{mysqld} with @code{--log-slow-queries}. @xref{Slow query log}.

If you find the text @code{mysqld restarted} in the error log file
(normally named @file{hostname.err}) you have probably found a query
that causes @code{mysqld} to fail.  If this happens you should check all
your tables with @code{myisamchk} (@pxref{MySQL Database Administration}),
and test the queries in the MySQL log files to see if one doesn't
work.  If you find such a query, try first upgrading to the newest
MySQL version.  If this doesn't help and you can't find anything
in the @code{mysql} mail archive, you should report the bug to
@email{mysql@@lists.mysql.com}.  Links to mail archives are available
online at @uref{http://lists.mysql.com/}.

If you have started @code{mysqld} with @code{myisam-recover},
MySQL will automatically check and try to repair @code{MyISAM}
tables if they are marked as 'not closed properly' or 'crashed'.  If
this happens, MySQL will write an entry in the
@code{hostname.err} file @code{'Warning: Checking table ...'} which is
followed by @code{Warning: Repairing table} if the table needs to be
repaired.  If you get a lot of these errors, without @code{mysqld} having
died unexpectedly just before, then something is wrong and needs to
be investigated further. @xref{Command-line options}.

It's of course not a good sign if @code{mysqld} did died unexpectedly,
but in this case one shouldn't investigate the @code{Checking table...}
messages but instead try to find out why @code{mysqld} died.


@node Reproduceable test case,  , Using log files, Debugging server
@appendixsubsec Making a Test Case When You Experience Table Corruption

If you get corrupted tables or if @code{mysqld} always fails after some
update commands, you can test if this bug is reproducible by doing the
following:

@itemize @bullet
@item
Take down the MySQL daemon (with @code{mysqladmin shutdown}).
@item
Make a backup of the tables (to guard against the very unlikely case that
the repair will do something bad).
@item
Check all tables with @code{myisamchk -s database/*.MYI}.  Repair any
wrong tables with @code{myisamchk -r database/table.MYI}.
@item
Make a second backup of the tables.
@item
Remove (or move away) any old log files from the MySQL data
directory if you need more space.
@item
Start @code{mysqld} with @code{--log-bin}. @xref{Binary log}.
If you want to find a query that crashes @code{mysqld}, you should use
@code{--log --log-bin}.
@item
When you have gotten a crashed table, stop the @code{mysqld server}.
@item
Restore the backup.
@item
Restart the @code{mysqld} server @strong{without} @code{--log-bin}
@item
Re-execute the commands with @code{mysqlbinlog update-log-file | mysql}.
The update log is saved in the MySQL database directory with
the name @code{hostname-bin.#}.
@item
If the tables are corrupted again or you can get @code{mysqld} to die with the
above command, you have found reproducible bug that should be easy to
fix! FTP the tables and the binary log to
@uref{ftp://support.mysql.com/pub/mysql/secret/} and send a mail to
@email{bugs@@lists.mysql.com} or (if you are a support customer) to
@email{support@@mysql.com} about the problem and the MySQL team
will fix it as soon as possible.
@end itemize

You can also use the script @code{mysql_find_rows} to just execute some of the
update statements if you want to narrow down the problem.


@node Debugging client, The DBUG package, Debugging server, Porting
@appendixsec Debugging a MySQL client

@cindex debugging, client
@cindex clients, debugging

To be able to debug a MySQL client with the integrated debug package,
you should configure MySQL with @code{--with-debug} or
@code{--with-debug=full}.  @xref{configure options}.

@tindex MYSQL_DEBUG environment variable
@tindex Environment variable, MYSQL_DEBUG
Before running a client, you should set the @code{MYSQL_DEBUG} environment
variable:

@example
shell> MYSQL_DEBUG=d:t:O,/tmp/client.trace
shell> export MYSQL_DEBUG
@end example

This causes clients to generate a trace file in @file{/tmp/client.trace}.

If you have problems with your own client code, you should attempt to
connect to the server and run your query using a client that is known to
work.  Do this by running @code{mysql} in debugging mode (assuming you
have compiled MySQL with debugging on):

@example
shell> mysql --debug=d:t:O,/tmp/client.trace
@end example

This will provide useful information in case you mail a bug report.
@xref{Bug reports}.

If your client crashes at some 'legal' looking code, you should check
that your @file{mysql.h} include file matches your mysql library file.
A very common mistake is to use an old @file{mysql.h} file from an old
MySQL installation with new MySQL library.


@node The DBUG package, Locking methods, Debugging client, Porting
@appendixsec The DBUG Package

@cindex DBUG package

The MySQL server and most MySQL clients are compiled
with the DBUG package originally made by Fred Fish.  When one has configured
MySQL for debugging, this package makes it possible to get a trace
file of what the program is debugging. @xref{Making trace files}.

One uses the debug package by invoking the program with the
@code{--debug="..."} or the @code{-#...} option.

Most MySQL programs has a default debug string that will be
used if you don't specify an option to @code{--debug}.  The default
trace file is usually @code{/tmp/programname.trace} on Unix and
@code{\programname.trace} on Windows.

The debug control string is a sequence of colon separated fields
as follows:

@example
<field_1>:<field_2>:...:<field_N>
@end example

Each field consists of a mandatory flag character followed by
an optional "," and comma-separated list of modifiers:

@example
flag[,modifier,modifier,...,modifier]
@end example

The currently recognised flag characters are:

@multitable @columnfractions .05 .95
@item @strong{Flag} @tab @strong{Description}
@item d	@tab Enable output from DBUG_<N> macros for the current state. May be followed by a list of keywords which selects output only for the DBUG macros with that keyword. An empty list of keywords implies output for all macros.
@item D	@tab Delay after each debugger output line. The argument is the number of tenths of seconds to delay, subject to machine capabilities. That is, @code{-#D,20} is delay two seconds.
@item f	@tab Limit debugging and/or tracing, and profiling to the list of named functions.  Note that a null list will disable all functions.	The appropriate "d" or "t" flags must still be given, this flag only limits their actions if they are enabled.
@item F	@tab Identify the source file name for each line of debug or trace output.
@item i	@tab Identify the process with the pid or thread id for each line of debug or trace output.
@item g	@tab Enable profiling.  Create a file called 'dbugmon.out' containing information that can be used to profile the program.  May be followed by a list of keywords that select profiling only for the functions in that list.  A null list implies that all functions are considered.
@item L	@tab Identify the source file line number for each line of debug or trace output.
@item n	@tab Print the current function nesting depth for each line of debug or trace output.
@item N @tab Number each line of dbug output.
@item o @tab Redirect the debugger output stream to the specified file.  The default output is stderr.
@item O @tab As @code{O} but the file is really flushed between each write. When needed the file is closed and reopened between each write.
@item p @tab Limit debugger actions to specified processes. A process must be identified with the DBUG_PROCESS macro and match one in the list for debugger actions to occur.
@item P @tab Print the current process name for each line of debug or trace output.
@item r @tab When pushing a new state, do not inherit the previous state's function nesting level. Useful when the output is to start at the left margin.
@item S @tab Do function _sanity(_file_,_line_) at each debugged function until _sanity() returns something that differs from 0. (Mostly used with safemalloc to find memory leaks)
@item t @tab Enable function call/exit trace lines. May be followed by a list (containing only one modifier) giving a numeric maximum trace level, beyond which no output will occur for either debugging or tracing macros.  The default is a compile time option.
@end multitable

Some examples of debug control strings which might appear on a shell
command-line (the "-#" is typically used to introduce a control string
to an application program) are:

@example
-#d:t
-#d:f,main,subr1:F:L:t,20
-#d,input,output,files:n
-#d:t:i:O,\\mysqld.trace
@end example

In MySQL, common tags to print (with the @code{d} option) are:
@code{enter},@code{exit},@code{error},@code{warning},@code{info} and
@code{loop}.


@node Locking methods, RTS-threads, The DBUG package, Porting
@appendixsec Locking methods

@cindex locking methods
@cindex methods, locking

Currently MySQL only supports table locking for
@code{ISAM}/@code{MyISAM} and @code{HEAP} tables,
page-level locking for @code{BDB} tables and
row-level locking for @code{InnoDB} tables.
@xref{Internal locking}.
With @code{MyISAM} tables one can freely mix @code{INSERT} and
@code{SELECT} without locks (@code{Versioning}).

Starting in version 3.23.33, you can analyse the table lock contention
on your system by checking @code{Table_locks_waited} and
@code{Table_locks_immediate} environment variables.

To decide if you want to use a table type with row-level locking,
you will want to look at what the application does and what the
select/update pattern of the data is.

Pros for row locking:

@itemize @bullet
@item
Fewer lock conflicts when accessing different rows in many threads.
@item
Fewer changes for rollbacks.
@item
Makes it possible to lock a single row a long time.
@end itemize

Cons:

@itemize @bullet
@item
Takes more memory than page level or table locks.
@item
Is slower than page level or table locks when used on a big part of the table,
because one has to do many more locks.
@item
Is definitely much worse than other locks if you do often do @code{GROUP
BY} on a large part of the data or if one has to often scan the whole table.
@item
With higher level locks one can also more easily support locks of
different types to tune the application as the lock overhead is less
notable as for row level locks.
@end itemize

Table locks are superior to page level / row level locks in the
following cases:

@itemize @bullet
@item
Mostly reads
@item
Read and updates on strict keys; this is where one updates or deletes
a row that can be fetched with one key read:
@example
UPDATE table_name SET column=value WHERE unique_key#
DELETE FROM table_name WHERE unique_key=#
@end example
@item
@code{SELECT} combined with @code{INSERT} (and very few @code{UPDATE}'s
and @code{DELETE}'s.
@item
Many scans / @code{GROUP BY} on the whole table without any writers.
@end itemize

Other options than row / page level locking:

Versioning (like we use in MySQL for concurrent inserts) where
you can have one writer at the same time as many readers.  This means
that the database/table supports different views for the data depending
on when one started to access it. Other names for this are time travel,
copy on write or copy on demand.

Copy on demand is in many case much better than page or row level
locking; the worst case does, however, use much more memory than
when using normal locks.

Instead of using row level locks one can use application level locks
(like get_lock/release_lock in MySQL).  This works of course
only in well-behaved applications.

In many cases one can do an educated guess which locking type is best
for the application but generally it's very hard to say that a given
lock type is better than another; everything depends on the application
and different part of the application may require different lock types.

Here are some tips about locking in MySQL:

On web application most applications do lots of selects, very few
deletes, updates mainly on keys and inserts in some specific tables.
The base MySQL setup is very well tuned for this.

Concurrent users is not a problem if one doesn't mix updates and selects
that needs to examine many rows in the same table.

If one mixes inserts and deletes on the same table then @code{INSERT DELAYED}
may be of great help.

One can also use @code{LOCK TABLES} to speed up things (many updates within
a single lock is much faster than updates without locks).  Splitting
thing to different tables will also helps.

If you get speed problems with the table locks in MySQL, you
may be able to solve these by converting some of your tables to @code{InnoDB}
or @code{BDB} tables.
@xref{InnoDB}. @xref{BDB}.

The optimisation section in the manual covers a lot of different aspects of
how to tune ones application. @xref{Tips}.


@node RTS-threads, Thread packages, Locking methods, Porting
@appendixsec Comments about RTS threads

@cindex RTS-threads
@cindex threads, RTS

I have tried to use the RTS thread packages with MySQL but
stumbled on the following problems:

They use an old version of a lot of POSIX calls and it is very tedious to
make wrappers for all functions. I am inclined to think that it would
be easier to change the thread libraries to the newest POSIX
specification.

Some wrappers are already written. See @file{mysys/my_pthread.c} for more info.

At least the following should be changed:

@code{pthread_get_specific} should use one argument.
@code{sigwait} should take two arguments.
A lot of functions (at least @code{pthread_cond_wait},
@code{pthread_cond_timedwait})
should return the error code on error. Now they return -1 and set @code{errno}.

Another problem is that user-level threads use the @code{ALRM} signal and this
aborts a lot of functions (@code{read}, @code{write}, @code{open}...).
MySQL should do a retry on interrupt on all of these but it is
not that easy to verify it.

The biggest unsolved problem is the following:

To get thread-level alarms I changed @file{mysys/thr_alarm.c} to wait between
alarms with @code{pthread_cond_timedwait()}, but this aborts with error
@code{EINTR}.  I tried to debug the thread library as to why this happens,
but couldn't find any easy solution.

If someone wants to try MySQL with RTS threads I suggest the
following:

@itemize @bullet
@item
Change functions MySQL uses from the thread library to POSIX.
This shouldn't take that long.
@item
Compile all libraries with the @code{-DHAVE_rts_threads}.
@item
Compile @code{thr_alarm}.
@item
If there are some small differences in the implementation, they may be fixed
by changing @file{my_pthread.h} and @file{my_pthread.c}.
@item
Run @code{thr_alarm}. If it runs without any ``warning'', ``error'' or aborted
messages, you are on the right track. Here is a successful run on
Solaris:
@example
Main thread: 1
Thread 0 (5) started
Thread: 5  Waiting
process_alarm
Thread 1 (6) started
Thread: 6  Waiting
process_alarm
process_alarm
thread_alarm
Thread: 6  Slept for 1 (1) sec
Thread: 6  Waiting
process_alarm
process_alarm
thread_alarm
Thread: 6  Slept for 2 (2) sec
Thread: 6  Simulation of no alarm needed
Thread: 6  Slept for 0 (3) sec
Thread: 6  Waiting
process_alarm
process_alarm
thread_alarm
Thread: 6  Slept for 4 (4) sec
Thread: 6  Waiting
process_alarm
thread_alarm
Thread: 5  Slept for 10 (10) sec
Thread: 5  Waiting
process_alarm
process_alarm
thread_alarm
Thread: 6  Slept for 5 (5) sec
Thread: 6  Waiting
process_alarm
process_alarm

...
thread_alarm
Thread: 5  Slept for 0 (1) sec
end
@end example
@end itemize


@node Thread packages,  , RTS-threads, Porting
@appendixsec Differences between different thread packages

@cindex thread packages, differences between

MySQL is very dependent on the thread package used. So when
choosing a good platform for MySQL, the thread package is very
important.

There are at least three types of thread packages:

@itemize @bullet
@item
User threads in a single process. Thread switching is managed with
alarms and the threads library manages all non-thread-safe functions
with locks. Read, write and select operations are usually managed with a
thread-specific select that switches to another thread if the running
threads have to wait for data.  If the user thread packages are
integrated in the standard libs (FreeBSD and BSDI threads) the thread
package requires less overhead than thread packages that have to map all
unsafe calls (MIT-pthreads, FSU Pthreads and RTS threads).  In some
environments (for example, SCO), all system calls are thread-safe so the
mapping can be done very easily (FSU Pthreads on SCO).  Downside: All
mapped calls take a little time and it's quite tricky to be able to
handle all situations. There are usually also some system calls that are
not handled by the thread package (like MIT-pthreads and sockets). Thread
scheduling isn't always optimal.
@item
User threads in separate processes. Thread switching is done by the
kernel and all data are shared between threads.  The thread package
manages the standard thread calls to allow sharing data between threads.
LinuxThreads is using this method.  Downside: Lots of processes. Thread
creating is slow. If one thread dies the rest are usually left hanging
and you must kill them all before restarting. Thread switching is
somewhat expensive.
@item
Kernel threads. Thread switching is handled by the thread library or the
kernel and is very fast. Everything is done in one process, but on some
systems, @code{ps} may show the different threads. If one thread aborts, the
whole process aborts. Most system calls are thread-safe and should
require very little overhead.  Solaris, HP-UX, AIX and OSF/1 have kernel
threads.
@end itemize

In some systems kernel threads are managed by integrating user
level threads in the system libraries. In such cases, the thread
switching can only be done by the thread library and the kernel isn't
really ``thread aware''.




@node Environment variables, Regexp, Porting, Top
@appendix Environment Variables

@cindex environment variables, list of

Here is a list of all the environment variables that are used directly or
indirectly by MySQL. Most of these can also be found in other
places in this manual.

Note that any options on the command-line will take precedence over
values specified in configuration files and environment variables, and
values in configuration files take precedence over values in environment
variables.

In many cases it's preferable to use a configure file instead of environment
variables to modify the behavior of MySQL. @xref{Option files}.

@tindex CCX environment variable
@tindex Environment variable, CCX
@tindex CC environment variable
@tindex Environment variable, CC
@tindex CFLAGS environment variable
@tindex Environment variable, CFLAGS
@tindex CXXFLAGS environment variable
@tindex Environment variable, CXXFLAGS
@tindex DBI_USER environment variable
@tindex Environment variable, DBI_USER
@tindex DBI_TRACE environment variable
@tindex Environment variable, DBI_TRACE
@tindex HOME environment variable
@tindex Environment variable, HOME
@tindex LD_RUN_PATH environment variable
@tindex Environment variable, LD_RUN_PATH
@tindex MYSQL_DEBUG environment variable
@tindex Environment variable, MYSQL_DEBUG
@tindex MYSQL_HISTFILE environment variable
@tindex Environment variable, MYSQL_HISTFILE
@tindex MYSQL_HOST environment variable
@tindex Environment variable, MYSQL_HOST
@tindex MYSQL_PWD environment variable
@tindex Environment variable, MYSQL_PWD
@tindex MYSQL_TCP_PORT environment variable
@tindex Environment variable, MYSQL_TCP_PORT
@tindex MYSQL_UNIX_PORT environment variable
@tindex Environment variable, MYSQL_UNIX_PORT
@tindex PATH environment variable
@tindex Environment variable, PATH
@tindex TMPDIR environment variable
@tindex Environment variable, TMPDIR
@tindex TZ environment variable
@tindex Environment variable, TZ
@tindex UMASK_DIR environment variable
@tindex Environment variable, UMASK_DIR
@tindex UMASK environment variable
@tindex Environment variable, UMASK
@tindex USER environment variable
@tindex Environment variable, USER

@multitable @columnfractions .25 .70
@item @strong{Variable} @tab @strong{Description}
@item @code{CCX}  @tab Set this to your C++ compiler when running configure.
@item @code{CC}  @tab Set this to your C compiler when running configure.
@item @code{CFLAGS} @tab Flags for your C compiler when running configure.
@item @code{CXXFLAGS} @tab Flags for your C++ compiler when running configure.
@item @code{DBI_USER} @tab The default user name for Perl DBI.
@item @code{DBI_TRACE} @tab Used when tracing Perl DBI.
@item @code{HOME} @tab The default path for the @code{mysql} history file is @file{$HOME/.mysql_history}.
@item @code{LD_RUN_PATH} @tab Used to specify where your @code{libmysqlclient.so} is.
@item @code{MYSQL_DEBUG} @tab Debug-trace options when debugging.
@item @code{MYSQL_HISTFILE} @tab The path to the @code{mysql} history file.
@item @code{MYSQL_HOST} @tab Default host name used by the @code{mysql} command-line prompt.
@item @code{MYSQL_PWD} @tab The default password when connecting to @code{mysqld}. Note that use of this is insecure!
@item @code{MYSQL_TCP_PORT} @tab The default TCP/IP port.
@item @code{MYSQL_UNIX_PORT} @tab The default socket; used for connections to @code{localhost}.
@item @code{PATH} @tab Used by the shell to finds the MySQL programs.
@item @code{TMPDIR} @tab The directory where temporary tables/files are created.
@item @code{TZ} @tab This should be set to your local time zone. @xref{Timezone problems}.
@item @code{UMASK_DIR} @tab The user-directory creation mask when creating directories. Note that this is ANDed with @code{UMASK}!
@item @code{UMASK} @tab The user-file creation mask when creating files.
@item @code{USER} @tab The default user on Windows to use when connecting to @code{mysqld}.
@end multitable




@node Regexp, GPL license, Environment variables, Top
@appendix MySQL Regular Expressions

@cindex regex
@cindex regular expression syntax, described
@cindex syntax, regular expression

A regular expression (regex) is a powerful way of specifying a complex search.

MySQL uses Henry Spencer's implementation of regular
expressions, which is aimed at conformance with POSIX
1003.2. MySQL uses the extended version.

This is a simplistic reference that skips the details. To get more exact
information, see Henry Spencer's @code{regex(7)} manual page that is
included in the source distribution. @xref{Credits}.

A regular expression describes a set of strings. The simplest regexp is
one that has no special characters in it. For example, the regexp
@code{hello} matches @code{hello} and nothing else.

Non-trivial regular expressions use certain special constructs so that
they can match more than one string. For example, the regexp
@code{hello|word} matches either the string @code{hello} or the string
@code{word}.

As a more complex example, the regexp @code{B[an]*s} matches any of the
strings @code{Bananas}, @code{Baaaaas}, @code{Bs}, and any other string
starting with a @code{B}, ending with an @code{s}, and containing any
number of @code{a} or @code{n} characters in between.

A regular expression may use any of the following special
characters/constructs:
@table @code
@item ^
Match the beginning of a string.
@example
mysql> SELECT "fo\nfo" REGEXP "^fo$";           -> 0
mysql> SELECT "fofo" REGEXP "^fo";              -> 1
@end example
@item $
Match the end of a string.
@example
mysql> SELECT "fo\no" REGEXP "^fo\no$";         -> 1
mysql> SELECT "fo\no" REGEXP "^fo$";            -> 0
@end example
@item .
Match any character (including newline).
@example
mysql> SELECT "fofo" REGEXP "^f.*";             -> 1
mysql> SELECT "fo\nfo" REGEXP "^f.*";           -> 1
@end example
@item a*
Match any sequence of zero or more @code{a} characters.
@example
mysql> SELECT "Ban" REGEXP "^Ba*n";             -> 1
mysql> SELECT "Baaan" REGEXP "^Ba*n";           -> 1
mysql> SELECT "Bn" REGEXP "^Ba*n";              -> 1
@end example
@item a+
Match any sequence of one or more @code{a} characters.
@example
mysql> SELECT "Ban" REGEXP "^Ba+n";             -> 1
mysql> SELECT "Bn" REGEXP "^Ba+n";              -> 0
@end example
@item a?
Match either zero or one @code{a} character.
@example
mysql> SELECT "Bn" REGEXP "^Ba?n";              -> 1
mysql> SELECT "Ban" REGEXP "^Ba?n";             -> 1
mysql> SELECT "Baan" REGEXP "^Ba?n";            -> 0
@end example
@item de|abc
Match either of the sequences @code{de} or @code{abc}.
@example
mysql> SELECT "pi" REGEXP "pi|apa";             -> 1
mysql> SELECT "axe" REGEXP "pi|apa";            -> 0
mysql> SELECT "apa" REGEXP "pi|apa";            -> 1
mysql> SELECT "apa" REGEXP "^(pi|apa)$";        -> 1
mysql> SELECT "pi" REGEXP "^(pi|apa)$";         -> 1
mysql> SELECT "pix" REGEXP "^(pi|apa)$";        -> 0
@end example
@item (abc)*
Match zero or more instances of the sequence @code{abc}.
@example
mysql> SELECT "pi" REGEXP "^(pi)*$";            -> 1
mysql> SELECT "pip" REGEXP "^(pi)*$";           -> 0
mysql> SELECT "pipi" REGEXP "^(pi)*$";          -> 1
@end example
@item @{1@}
@itemx @{2,3@}
The is a more general way of writing regexps that match many
occurrences of the previous atom.
@table @code
@item a*
Can be written as @code{a@{0,@}}.
@item a+
Can be written as @code{a@{1,@}}.
@item a?
Can be written as @code{a@{0,1@}}.
@end table
To be more precise, an atom followed by a bound containing one integer
@code{i} and no comma matches a sequence of exactly @code{i} matches of
the atom. An atom followed by a bound containing one integer @code{i}
and a comma matches a sequence of @code{i} or more matches of the atom.
An atom followed by a bound containing two integers @code{i} and
@code{j} matches a sequence of @code{i} through @code{j} (inclusive)
matches of the atom.

Both arguments must be in the range from @code{0} to @code{RE_DUP_MAX}
(default 255), inclusive.  If there are two arguments, the second must be
greater than or equal to the first.
@item [a-dX]
@itemx [^a-dX]
Matches
any character which is (or is not, if ^ is used) either @code{a}, @code{b},
@code{c}, @code{d} or @code{X}. To include a literal @code{]} character,
it must immediately follow the opening bracket @code{[}.  To include a
literal @code{-} character, it must be written first or last. So
@code{[0-9]} matches any decimal digit. Any character that does not have
a defined meaning inside a @code{[]} pair has no special meaning and
matches only itself.
@example
mysql> SELECT "aXbc" REGEXP "[a-dXYZ]";         -> 1
mysql> SELECT "aXbc" REGEXP "^[a-dXYZ]$";       -> 0
mysql> SELECT "aXbc" REGEXP "^[a-dXYZ]+$";      -> 1
mysql> SELECT "aXbc" REGEXP "^[^a-dXYZ]+$";     -> 0
mysql> SELECT "gheis" REGEXP "^[^a-dXYZ]+$";    -> 1
mysql> SELECT "gheisa" REGEXP "^[^a-dXYZ]+$";   -> 0
@end example
@item [[.characters.]]
The sequence of characters of that collating element. The sequence is a
single element of the bracket expression's list. A bracket expression
containing a multi-character collating element can thus match more than
one character, for example, if the collating sequence includes a @code{ch}
collating element, then the regular expression @code{[[.ch.]]*c} matches the
first five characters of @code{chchcc}.

@item [=character_class=]
An equivalence class, standing for the sequences of characters of all
collating elements equivalent to that one, including itself.

For example, if @code{o} and @code{(+)} are the members of an
equivalence class, then @code{[[=o=]]}, @code{[[=(+)=]]}, and
@code{[o(+)]} are all synonymous. An equivalence class may not be an
endpoint of a range.

@item [:character_class:]
Within a bracket expression, the name of a character class enclosed in
@code{[:} and @code{:]} stands for the list of all characters belonging
to that class. Standard character class names are:

@multitable @columnfractions .10 .10 .10
@item @strong{Name} @tab @strong{Name} @tab @strong{Name}
@item alnum @tab digit @tab punct
@item alpha @tab graph @tab space
@item blank @tab lower @tab upper
@item cntrl @tab print @tab xdigit
@end multitable

These stand for the character classes defined in the @code{ctype(3)} manual
page. A locale may provide others. A character class may not be used as an
endpoint of a range.
@example
mysql> SELECT "justalnums" REGEXP "[[:alnum:]]+";       -> 1
mysql> SELECT "!!" REGEXP "[[:alnum:]]+";               -> 0
@end example

@item [[:<:]]
@itemx [[:>:]]
These match the null string at the beginning and end of a word
respectively.  A word is defined as a sequence of word characters which
is neither preceded nor followed by word characters. A word character is
an alnum character (as defined by @code{ctype(3)}) or an underscore
(@code{_}).
@example
mysql> SELECT "a word a" REGEXP "[[:<:]]word[[:>:]]";      -> 1
mysql> SELECT "a xword a" REGEXP "[[:<:]]word[[:>:]]";     -> 0
@end example
@end table

@example
mysql> SELECT "weeknights" REGEXP "^(wee|week)(knights|nights)$"; -> 1
@end example




@node GPL license, LGPL license, Regexp, Top
@appendix GNU General Public License

@cindex GPL, General Public License
@cindex GPL, GNU General Public License

@center Version 2, June 1991

@display
Copyright @copyright{} 1989, 1991 Free Software Foundation, Inc.
59 Temple Place - Suite 330, Boston, MA  02111-1307, USA

Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
@end display

@appendixsec Preamble

  The licenses for most software are designed to take away your
freedom to share and change it.  By contrast, the GNU General Public
License is intended to guarantee your freedom to share and change free
software---to make sure the software is free for all its users.  This
General Public License applies to most of the Free Software
Foundation's software and to any other program whose authors commit to
using it.  (Some other Free Software Foundation software is covered by
the GNU Library General Public License instead.)  You can apply it to
your programs, too.

  When we speak of free software, we are referring to freedom, not
price.  Our General Public Licenses are designed to make sure that you
have the freedom to distribute copies of free software (and charge for
this service if you wish), that you receive source code or can get it
if you want it, that you can change the software or use pieces of it
in new free programs; and that you know you can do these things.

  To protect your rights, we need to make restrictions that forbid
anyone to deny you these rights or to ask you to surrender the rights.
These restrictions translate to certain responsibilities for you if you
distribute copies of the software, or if you modify it.

  For example, if you distribute copies of such a program, whether
gratis or for a fee, you must give the recipients all the rights that
you have.  You must make sure that they, too, receive or can get the
source code.  And you must show them these terms so they know their
rights.

  We protect your rights with two steps: (1) copyright the software, and
(2) offer you this license which gives you legal permission to copy,
distribute and/or modify the software.

  Also, for each author's protection and ours, we want to make certain
that everyone understands that there is no warranty for this free
software.  If the software is modified by someone else and passed on, we
want its recipients to know that what they have is not the original, so
that any problems introduced by others will not reflect on the original
authors' reputations.

  Finally, any free program is threatened constantly by software
patents.  We wish to avoid the danger that redistributors of a free
program will individually obtain patent licenses, in effect making the
program proprietary.  To prevent this, we have made it clear that any
patent must be licensed for everyone's free use or not licensed at all.

  The precise terms and conditions for copying, distribution and
modification follow.

@iftex
@appendixsec TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
@end iftex
@ifinfo
@center GNU GENERAL PUBLIC LICENSE
@center TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
@end ifinfo

@enumerate 0
@item
This License applies to any program or other work which contains
a notice placed by the copyright holder saying it may be distributed
under the terms of this General Public License.  The ``Program'', below,
refers to any such program or work, and a ``work based on the Program''
means either the Program or any derivative work under copyright law:
that is to say, a work containing the Program or a portion of it,
either verbatim or with modifications and/or translated into another
language.  (Hereinafter, translation is included without limitation in
the term ``modification''.)  Each licensee is addressed as ``you''.

Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope.  The act of
running the Program is not restricted, and the output from the Program
is covered only if its contents constitute a work based on the
Program (independent of having been made by running the Program).
Whether that is true depends on what the Program does.

@item
You may copy and distribute verbatim copies of the Program's
source code as you receive it, in any medium, provided that you
conspicuously and appropriately publish on each copy an appropriate
copyright notice and disclaimer of warranty; keep intact all the
notices that refer to this License and to the absence of any warranty;
and give any other recipients of the Program a copy of this License
along with the Program.

You may charge a fee for the physical act of transferring a copy, and
you may at your option offer warranty protection in exchange for a fee.

@item
You may modify your copy or copies of the Program or any portion
of it, thus forming a work based on the Program, and copy and
distribute such modifications or work under the terms of Section 1
above, provided that you also meet all of these conditions:

@enumerate a
@item
You must cause the modified files to carry prominent notices
stating that you changed the files and the date of any change.

@item
You must cause any work that you distribute or publish, that in
whole or in part contains or is derived from the Program or any
part thereof, to be licensed as a whole at no charge to all third
parties under the terms of this License.

@item
If the modified program normally reads commands interactively
when run, you must cause it, when started running for such
interactive use in the most ordinary way, to print or display an
announcement including an appropriate copyright notice and a
notice that there is no warranty (or else, saying that you provide
a warranty) and that users may redistribute the program under
these conditions, and telling the user how to view a copy of this
License.  (Exception: if the Program itself is interactive but
does not normally print such an announcement, your work based on
the Program is not required to print an announcement.)
@end enumerate

These requirements apply to the modified work as a whole.  If
identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works.  But when you
distribute the same sections as part of a whole which is a work based
on the Program, the distribution of the whole must be on the terms of
this License, whose permissions for other licensees extend to the
entire whole, and thus to each and every part regardless of who wrote it.

Thus, it is not the intent of this section to claim rights or contest
your rights to work written entirely by you; rather, the intent is to
exercise the right to control the distribution of derivative or
collective works based on the Program.

In addition, mere aggregation of another work not based on the Program
with the Program (or with a work based on the Program) on a volume of
a storage or distribution medium does not bring the other work under
the scope of this License.

@item
You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:

@enumerate a
@item
Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections
1 and 2 above on a medium customarily used for software interchange; or,

@item
Accompany it with a written offer, valid for at least three
years, to give any third-party, for a charge no more than your
cost of physically performing source distribution, a complete
machine-readable copy of the corresponding source code, to be
distributed under the terms of Sections 1 and 2 above on a medium
customarily used for software interchange; or,

@item
Accompany it with the information you received as to the offer
to distribute corresponding source code.  (This alternative is
allowed only for noncommercial distribution and only if you
received the program in object code or executable form with such
an offer, in accord with Subsection b above.)
@end enumerate

The source code for a work means the preferred form of the work for
making modifications to it.  For an executable work, complete source
code means all the source code for all modules it contains, plus any
associated interface definition files, plus the scripts used to
control compilation and installation of the executable.  However, as a
special exception, the source code distributed need not include
anything that is normally distributed (in either source or binary
form) with the major components (compiler, kernel, and so on) of the
operating system on which the executable runs, unless that component
itself accompanies the executable.

If distribution of executable or object code is made by offering
access to copy from a designated place, then offering equivalent
access to copy the source code from the same place counts as
distribution of the source code, even though third parties are not
compelled to copy the source along with the object code.

@item
You may not copy, modify, sublicense, or distribute the Program
except as expressly provided under this License.  Any attempt
otherwise to copy, modify, sublicense or distribute the Program is
void, and will automatically terminate your rights under this License.
However, parties who have received copies, or rights, from you under
this License will not have their licenses terminated so long as such
parties remain in full compliance.

@item
You are not required to accept this License, since you have not
signed it.  However, nothing else grants you permission to modify or
distribute the Program or its derivative works.  These actions are
prohibited by law if you do not accept this License.  Therefore, by
modifying or distributing the Program (or any work based on the
Program), you indicate your acceptance of this License to do so, and
all its terms and conditions for copying, distributing or modifying
the Program or works based on it.

@item
Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions.  You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties to
this License.

@item
If, as a consequence of a court judgment or allegation of patent
infringement or for any other reason (not limited to patent issues),
conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License.  If you cannot
distribute so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you
may not distribute the Program at all.  For example, if a patent
license would not permit royalty-free redistribution of the Program by
all those who receive copies directly or indirectly through you, then
the only way you could satisfy both it and this License would be to
refrain entirely from distribution of the Program.

If any portion of this section is held invalid or unenforceable under
any particular circumstance, the balance of the section is intended to
apply and the section as a whole is intended to apply in other
circumstances.

It is not the purpose of this section to induce you to infringe any
patents or other property right claims or to contest validity of any
such claims; this section has the sole purpose of protecting the
integrity of the free software distribution system, which is
implemented by public license practices.  Many people have made
generous contributions to the wide range of software distributed
through that system in reliance on consistent application of that
system; it is up to the author/donor to decide if he or she is willing
to distribute software through any other system and a licensee cannot
impose that choice.

This section is intended to make thoroughly clear what is believed to
be a consequence of the rest of this License.

@item
If the distribution and/or use of the Program is restricted in
certain countries either by patents or by copyrighted interfaces, the
original copyright holder who places the Program under this License
may add an explicit geographical distribution limitation excluding
those countries, so that distribution is permitted only in or among
countries not thus excluded.  In such case, this License incorporates
the limitation as if written in the body of this License.

@item
The Free Software Foundation may publish revised and/or new versions
of the General Public License from time to time.  Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.

Each version is given a distinguishing version number.  If the Program
specifies a version number of this License which applies to it and ``any
later version'', you have the option of following the terms and conditions
either of that version or of any later version published by the Free
Software Foundation.  If the Program does not specify a version number of
this License, you may choose any version ever published by the Free Software
Foundation.

@item
If you wish to incorporate parts of the Program into other free
programs whose distribution conditions are different, write to the author
to ask for permission.  For software which is copyrighted by the Free
Software Foundation, write to the Free Software Foundation; we sometimes
make exceptions for this.  Our decision will be guided by the two goals
of preserving the free status of all derivatives of our free software and
of promoting the sharing and reuse of software generally.

@iftex
@heading NO WARRANTY
@end iftex
@ifinfo
@center NO WARRANTY
@end ifinfo

@item
BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
PROVIDE THE PROGRAM ``AS IS'' WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS
TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE
PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
REPAIR OR CORRECTION.

@item
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
POSSIBILITY OF SUCH DAMAGES.
@end enumerate

@iftex
@heading END OF TERMS AND CONDITIONS
@end iftex
@ifinfo
@center END OF TERMS AND CONDITIONS
@end ifinfo

@page
@appendixsec How to Apply These Terms to Your New Programs

  If you develop a new program, and you want it to be of the greatest
possible use to the public, the best way to achieve this is to make it
free software which everyone can redistribute and change under these terms.

  To do so, attach the following notices to the program.  It is safest
to attach them to the start of each source file to most effectively
convey the exclusion of warranty; and each file should have at least
the ``copyright'' line and a pointer to where the full notice is found.

@smallexample
@var{one line to give the program's name and a brief idea of what it does.}
Copyright (C) @var{yyyy}  @var{name of author}

This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
@end smallexample

Also add information on how to contact you by electronic and paper mail.

If the program is interactive, make it output a short notice like this
when it starts in an interactive mode:

@smallexample
Gnomovision version 69, Copyright (C) 19@var{yy} @var{name of author}
Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
This is free software, and you are welcome to redistribute it
under certain conditions; type `show c' for details.
@end smallexample

The hypothetical commands @samp{show w} and @samp{show c} should show
the appropriate parts of the General Public License.  Of course, the
commands you use may be called something other than @samp{show w} and
@samp{show c}; they could even be mouse-clicks or menu items---whatever
suits your program.

You should also get your employer (if you work as a programmer) or your
school, if any, to sign a ``copyright disclaimer'' for the program, if
necessary.  Here is a sample; alter the names:

@example
Yoyodyne, Inc., hereby disclaims all copyright interest in the program
`Gnomovision' (which makes passes at compilers) written by James Hacker.

@var{signature of Ty Coon}, 1 April 1989
Ty Coon, President of Vice
@end example

This General Public License does not permit incorporating your program into
proprietary programs.  If your program is a subroutine library, you may
consider it more useful to permit linking proprietary applications with the
library.  If this is what you want to do, use the GNU Library General
Public License instead of this License.

@page




@node LGPL license, Function Index, GPL license, Top
@appendix GNU Lesser General Public License

@cindex LGPL, Library General Public License
@cindex LGPL, GNU Library General Public License
@cindex LGPL, Lesser General Public License
@cindex LGPL, GNU Lesser General Public License

@center Version 2.1, February 1999

@display
Copyright @copyright{} 1991, 1999 Free Software Foundation, Inc.
59 Temple Place -- Suite 330, Boston, MA 02111-1307, USA

Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.

[This is the first released version of the Lesser GPL.  It also counts
as the successor of the GNU Library Public License, version 2, hence the
version number 2.1.]
@end display

@appendixsec Preamble

  The licenses for most software are designed to take away your
freedom to share and change it.  By contrast, the GNU General Public
Licenses are intended to guarantee your freedom to share and change
free software---to make sure the software is free for all its users.

  This license, the Lesser General Public License, applies to some
specially designated software---typically libraries---of the Free
Software Foundation and other authors who decide to use it.  You can use
it too, but we suggest you first think carefully about whether this
license or the ordinary General Public License is the better strategy to
use in any particular case, based on the explanations below.

  When we speak of free software, we are referring to freedom of use,
not price.  Our General Public Licenses are designed to make sure that
you have the freedom to distribute copies of free software (and charge
for this service if you wish); that you receive source code or can get
it if you want it; that you can change the software and use pieces of it
in new free programs; and that you are informed that you can do these
things.

  To protect your rights, we need to make restrictions that forbid
distributors to deny you these rights or to ask you to surrender these
rights.  These restrictions translate to certain responsibilities for
you if you distribute copies of the library or if you modify it.

  For example, if you distribute copies of the library, whether gratis
or for a fee, you must give the recipients all the rights that we gave
you.  You must make sure that they, too, receive or can get the source
code.  If you link other code with the library, you must provide
complete object files to the recipients, so that they can relink them
with the library after making changes to the library and recompiling
it.  And you must show them these terms so they know their rights.

  We protect your rights with a two-step method: (1) we copyright the
library, and (2) we offer you this license, which gives you legal
permission to copy, distribute and/or modify the library.

  To protect each distributor, we want to make it very clear that
there is no warranty for the free library.  Also, if the library is
modified by someone else and passed on, the recipients should know
that what they have is not the original version, so that the original
author's reputation will not be affected by problems that might be
introduced by others.

  Finally, software patents pose a constant threat to the existence of
any free program.  We wish to make sure that a company cannot
effectively restrict the users of a free program by obtaining a
restrictive license from a patent holder.  Therefore, we insist that
any patent license obtained for a version of the library must be
consistent with the full freedom of use specified in this license.

  Most GNU software, including some libraries, is covered by the
ordinary GNU General Public License.  This license, the GNU Lesser
General Public License, applies to certain designated libraries, and
is quite different from the ordinary General Public License.  We use
this license for certain libraries in order to permit linking those
libraries into non-free programs.

  When a program is linked with a library, whether statically or using
a shared library, the combination of the two is legally speaking a
combined work, a derivative of the original library.  The ordinary
General Public License therefore permits such linking only if the
entire combination fits its criteria of freedom.  The Lesser General
Public License permits more lax criteria for linking other code with
the library.

  We call this license the @dfn{Lesser} General Public License because it
does @emph{Less} to protect the user's freedom than the ordinary General
Public License.  It also provides other free software developers Less
of an advantage over competing non-free programs.  These disadvantages
are the reason we use the ordinary General Public License for many
libraries.  However, the Lesser license provides advantages in certain
special circumstances.

  For example, on rare occasions, there may be a special need to
encourage the widest possible use of a certain library, so that it becomes
a de-facto standard.  To achieve this, non-free programs must be
allowed to use the library.  A more frequent case is that a free
library does the same job as widely used non-free libraries.  In this
case, there is little to gain by limiting the free library to free
software only, so we use the Lesser General Public License.

  In other cases, permission to use a particular library in non-free
programs enables a greater number of people to use a large body of
free software.  For example, permission to use the GNU C Library in
non-free programs enables many more people to use the whole GNU
operating system, as well as its variant, the GNU/Linux operating
system.

  Although the Lesser General Public License is Less protective of the
users' freedom, it does ensure that the user of a program that is
linked with the Library has the freedom and the wherewithal to run
that program using a modified version of the Library.

  The precise terms and conditions for copying, distribution and
modification follow.  Pay close attention to the difference between a
``work based on the library'' and a ``work that uses the library''.  The
former contains code derived from the library, whereas the latter must
be combined with the library in order to run.

@iftex
@appendixsec TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
@end iftex
@ifinfo
@center GNU LESSER GENERAL PUBLIC LICENSE
@center TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
@end ifinfo

@enumerate 0
@item
This License Agreement applies to any software library or other program
which contains a notice placed by the copyright holder or other
authorized party saying it may be distributed under the terms of this
Lesser General Public License (also called ``this License'').  Each
licensee is addressed as ``you''.

  A ``library'' means a collection of software functions and/or data
prepared so as to be conveniently linked with application programs
(which use some of those functions and data) to form executables.

  The ``Library'', below, refers to any such software library or work
which has been distributed under these terms.  A ``work based on the
Library'' means either the Library or any derivative work under
copyright law: that is to say, a work containing the Library or a
portion of it, either verbatim or with modifications and/or translated
straightforwardly into another language.  (Hereinafter, translation is
included without limitation in the term ``modification''.)

  ``Source code'' for a work means the preferred form of the work for
making modifications to it.  For a library, complete source code means
all the source code for all modules it contains, plus any associated
interface definition files, plus the scripts used to control compilation
and installation of the library.

  Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope.  The act of
running a program using the Library is not restricted, and output from
such a program is covered only if its contents constitute a work based
on the Library (independent of the use of the Library in a tool for
writing it).  Whether that is true depends on what the Library does
and what the program that uses the Library does.

@item
You may copy and distribute verbatim copies of the Library's
complete source code as you receive it, in any medium, provided that
you conspicuously and appropriately publish on each copy an
appropriate copyright notice and disclaimer of warranty; keep intact
all the notices that refer to this License and to the absence of any
warranty; and distribute a copy of this License along with the
Library.

  You may charge a fee for the physical act of transferring a copy,
and you may at your option offer warranty protection in exchange for a
fee.

@item
You may modify your copy or copies of the Library or any portion
of it, thus forming a work based on the Library, and copy and
distribute such modifications or work under the terms of Section 1
above, provided that you also meet all of these conditions:

@enumerate a
@item
The modified work must itself be a software library.

@item
You must cause the files modified to carry prominent notices
stating that you changed the files and the date of any change.

@item
You must cause the whole of the work to be licensed at no
charge to all third parties under the terms of this License.

@item
If a facility in the modified Library refers to a function or a
table of data to be supplied by an application program that uses
the facility, other than as an argument passed when the facility
is invoked, then you must make a good faith effort to ensure that,
in the event an application does not supply such function or
table, the facility still operates, and performs whatever part of
its purpose remains meaningful.

(For example, a function in a library to compute square roots has
a purpose that is entirely well-defined independent of the
application.  Therefore, Subsection 2d requires that any
application-supplied function or table used by this function must
be optional: if the application does not supply it, the square
root function must still compute square roots.)
@end enumerate

These requirements apply to the modified work as a whole.  If
identifiable sections of that work are not derived from the Library,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works.  But when you
distribute the same sections as part of a whole which is a work based
on the Library, the distribution of the whole must be on the terms of
this License, whose permissions for other licensees extend to the
entire whole, and thus to each and every part regardless of who wrote
it.

Thus, it is not the intent of this section to claim rights or contest
your rights to work written entirely by you; rather, the intent is to
exercise the right to control the distribution of derivative or
collective works based on the Library.

In addition, mere aggregation of another work not based on the Library
with the Library (or with a work based on the Library) on a volume of
a storage or distribution medium does not bring the other work under
the scope of this License.

@item
You may opt to apply the terms of the ordinary GNU General Public
License instead of this License to a given copy of the Library.  To do
this, you must alter all the notices that refer to this License, so
that they refer to the ordinary GNU General Public License, version 2,
instead of to this License.  (If a newer version than version 2 of the
ordinary GNU General Public License has appeared, then you can specify
that version instead if you wish.)  Do not make any other change in
these notices.

  Once this change is made in a given copy, it is irreversible for
that copy, so the ordinary GNU General Public License applies to all
subsequent copies and derivative works made from that copy.

  This option is useful when you wish to copy part of the code of
the Library into a program that is not a library.

@item
You may copy and distribute the Library (or a portion or
derivative of it, under Section 2) in object code or executable form
under the terms of Sections 1 and 2 above provided that you accompany
it with the complete corresponding machine-readable source code, which
must be distributed under the terms of Sections 1 and 2 above on a
medium customarily used for software interchange.

  If distribution of object code is made by offering access to copy
from a designated place, then offering equivalent access to copy the
source code from the same place satisfies the requirement to
distribute the source code, even though third parties are not
compelled to copy the source along with the object code.

@item
A program that contains no derivative of any portion of the
Library, but is designed to work with the Library by being compiled or
linked with it, is called a ``work that uses the Library''.  Such a
work, in isolation, is not a derivative work of the Library, and
therefore falls outside the scope of this License.

  However, linking a ``work that uses the Library'' with the Library
creates an executable that is a derivative of the Library (because it
contains portions of the Library), rather than a ``work that uses the
library''.  The executable is therefore covered by this License.
Section 6 states terms for distribution of such executables.

  When a ``work that uses the Library'' uses material from a header file
that is part of the Library, the object code for the work may be a
derivative work of the Library even though the source code is not.
Whether this is true is especially significant if the work can be
linked without the Library, or if the work is itself a library.  The
threshold for this to be true is not precisely defined by law.

  If such an object file uses only numerical parameters, data
structure layouts and accessors, and small macros and small inline
functions (ten lines or less in length), then the use of the object
file is unrestricted, regardless of whether it is legally a derivative
work.  (Executables containing this object code plus portions of the
Library will still fall under Section 6.)

  Otherwise, if the work is a derivative of the Library, you may
distribute the object code for the work under the terms of Section 6.
Any executables containing that work also fall under Section 6,
whether they are linked directly with the Library itself.

@item
As an exception to the Sections above, you may also combine or
link a ``work that uses the Library'' with the Library to produce a
work containing portions of the Library, and distribute that work
under terms of your choice, provided that the terms permit
modification of the work for the customer's own use and reverse
engineering for debugging such modifications.

  You must give prominent notice with each copy of the work that the
Library is used in it and that the Library and its use are covered by
this License.  You must supply a copy of this License.  If the work
during execution displays copyright notices, you must include the
copyright notice for the Library among them, as well as a reference
directing the user to the copy of this License.  Also, you must do one
of these things:

@enumerate a
@item
Accompany the work with the complete corresponding
machine-readable source code for the Library including whatever
changes were used in the work (which must be distributed under
Sections 1 and 2 above); and, if the work is an executable linked
with the Library, with the complete machine-readable ``work that
uses the Library'', as object code and/or source code, so that the
user can modify the Library and then relink to produce a modified
executable containing the modified Library.  (It is understood
that the user who changes the contents of definitions files in the
Library will not necessarily be able to recompile the application
to use the modified definitions.)

@item
Use a suitable shared library mechanism for linking with the Library.  A
suitable mechanism is one that (1) uses at runtime a copy of the
library already present on the user's computer system, rather than
copying library functions into the executable, and (2) will operate
properly with a modified version of the library, if the user installs
one, as long as the modified version is interface-compatible with the
version that the work was made with.

@item
Accompany the work with a written offer, valid for at
least three years, to give the same user the materials
specified in Subsection 6a, above, for a charge no more
than the cost of performing this distribution.

@item
If distribution of the work is made by offering access to copy
from a designated place, offer equivalent access to copy the above
specified materials from the same place.

@item
Verify that the user has already received a copy of these
materials or that you have already sent this user a copy.
@end enumerate

  For an executable, the required form of the ``work that uses the
Library'' must include any data and utility programs needed for
reproducing the executable from it.  However, as a special exception,
the materials to be distributed need not include anything that is
normally distributed (in either source or binary form) with the major
components (compiler, kernel, and so on) of the operating system on
which the executable runs, unless that component itself accompanies the
executable.

  It may happen that this requirement contradicts the license
restrictions of other proprietary libraries that do not normally
accompany the operating system.  Such a contradiction means you cannot
use both them and the Library together in an executable that you
distribute.

@item
You may place library facilities that are a work based on the
Library side-by-side in a single library together with other library
facilities not covered by this License, and distribute such a combined
library, provided that the separate distribution of the work based on
the Library and of the other library facilities is otherwise
permitted, and provided that you do these two things:

@enumerate a
@item
Accompany the combined library with a copy of the same work
based on the Library, uncombined with any other library
facilities.  This must be distributed under the terms of the
Sections above.

@item
Give prominent notice with the combined library of the fact
that part of it is a work based on the Library, and explaining
where to find the accompanying uncombined form of the same work.
@end enumerate

@item
You may not copy, modify, sublicense, link with, or distribute
the Library except as expressly provided under this License.  Any
attempt otherwise to copy, modify, sublicense, link with, or
distribute the Library is void, and will automatically terminate your
rights under this License.  However, parties who have received copies,
or rights, from you under this License will not have their licenses
terminated so long as such parties remain in full compliance.

@item
You are not required to accept this License, since you have not
signed it.  However, nothing else grants you permission to modify or
distribute the Library or its derivative works.  These actions are
prohibited by law if you do not accept this License.  Therefore, by
modifying or distributing the Library (or any work based on the
Library), you indicate your acceptance of this License to do so, and
all its terms and conditions for copying, distributing or modifying
the Library or works based on it.

@item
Each time you redistribute the Library (or any work based on the
Library), the recipient automatically receives a license from the
original licensor to copy, distribute, link with or modify the Library
subject to these terms and conditions.  You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties with
this License.

@item
If, as a consequence of a court judgment or allegation of patent
infringement or for any other reason (not limited to patent issues),
conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License.  If you cannot
distribute so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you
may not distribute the Library at all.  For example, if a patent
license would not permit royalty-free redistribution of the Library by
all those who receive copies directly or indirectly through you, then
the only way you could satisfy both it and this License would be to
refrain entirely from distribution of the Library.

If any portion of this section is held invalid or unenforceable under any
particular circumstance, the balance of the section is intended to apply,
and the section as a whole is intended to apply in other circumstances.

It is not the purpose of this section to induce you to infringe any
patents or other property right claims or to contest validity of any
such claims; this section has the sole purpose of protecting the
integrity of the free software distribution system which is
implemented by public license practices.  Many people have made
generous contributions to the wide range of software distributed
through that system in reliance on consistent application of that
system; it is up to the author/donor to decide if he or she is willing
to distribute software through any other system and a licensee cannot
impose that choice.

This section is intended to make thoroughly clear what is believed to
be a consequence of the rest of this License.

@item
If the distribution and/or use of the Library is restricted in
certain countries either by patents or by copyrighted interfaces, the
original copyright holder who places the Library under this License may add
an explicit geographical distribution limitation excluding those countries,
so that distribution is permitted only in or among countries not thus
excluded.  In such case, this License incorporates the limitation as if
written in the body of this License.

@item
The Free Software Foundation may publish revised and/or new
versions of the Lesser General Public License from time to time.
Such new versions will be similar in spirit to the present version,
but may differ in detail to address new problems or concerns.

Each version is given a distinguishing version number.  If the Library
specifies a version number of this License which applies to it and
``any later version'', you have the option of following the terms and
conditions either of that version or of any later version published by
the Free Software Foundation.  If the Library does not specify a
license version number, you may choose any version ever published by
the Free Software Foundation.

@item
If you wish to incorporate parts of the Library into other free
programs whose distribution conditions are incompatible with these,
write to the author to ask for permission.  For software which is
copyrighted by the Free Software Foundation, write to the Free
Software Foundation; we sometimes make exceptions for this.  Our
decision will be guided by the two goals of preserving the free status
of all derivatives of our free software and of promoting the sharing
and reuse of software generally.

@iftex
@heading NO WARRANTY
@end iftex
@ifinfo
@center NO WARRANTY
@end ifinfo

@item
BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR
OTHER PARTIES PROVIDE THE LIBRARY ``AS IS'' WITHOUT WARRANTY OF ANY
KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE.  THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE
LIBRARY IS WITH YOU.  SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME
THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.

@item
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY
AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU
FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE
LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING
RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A
FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF
SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
DAMAGES.
@end enumerate

@iftex
@heading END OF TERMS AND CONDITIONS
@end iftex
@ifinfo
@center END OF TERMS AND CONDITIONS
@end ifinfo

@page
@appendixsec How to Apply These Terms to Your New Libraries

  If you develop a new library, and you want it to be of the greatest
possible use to the public, we recommend making it free software that
everyone can redistribute and change.  You can do so by permitting
redistribution under these terms (or, alternatively, under the terms of the
ordinary General Public License).

  To apply these terms, attach the following notices to the library.  It is
safest to attach them to the start of each source file to most effectively
convey the exclusion of warranty; and each file should have at least the
``copyright'' line and a pointer to where the full notice is found.

@smallexample
@var{one line to give the library's name and an idea of what it does.}
Copyright (C) @var{year}  @var{name of author}

This library is free software; you can redistribute it and/or modify it
under the terms of the GNU Lesser General Public License as published by
the Free Software Foundation; either version 2.1 of the License, or (at
your option) any later version.

This library is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307,
USA.
@end smallexample

Also add information on how to contact you by electronic and paper mail.

You should also get your employer (if you work as a programmer) or your
school, if any, to sign a ``copyright disclaimer'' for the library, if
necessary.  Here is a sample; alter the names:

@smallexample
Yoyodyne, Inc., hereby disclaims all copyright interest in the library
`Frob' (a library for tweaking knobs) written by James Random Hacker.

@var{signature of Ty Coon}, 1 April 1990
Ty Coon, President of Vice
@end smallexample

That's all there is to it!




@node Function Index, Concept Index, LGPL license, Top
@unnumbered SQL command, type and function index

@printindex fn

@page


@node Concept Index,  , Function Index, Top
@unnumbered Concept Index

@printindex cp





@bye