| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
| |
The shutdown_sync was actually interfering with a clean shutdown
initiated by the index processes themselves when they received a 'DOWN'
message from their monitoring of the DB process. The explicit 'delete'
cast message in this patch is very probably redundant, but it makes the
intent clear. The important bit is to let the index process terminate
*and* get confirmation of the termination before attempting to nuke the
directory.
COUCHDB-1364
|
|
|
|
|
| |
- introduced in COUCHDB-1338
- allows building cleanly on Windows without cURL
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently calls to getRow() cause the HTTP headers to be sent immediately back
to the client. This happens even if an error is thrown after the getRow(), but
before any send(...) or start(...). Worse, if a list throws an exception an
extra, invalid header is sent to the client (resulting in various bad
behavior).
Erlang list handling will now wait until data has been sent BEFORE sending the
HTTP headers to the client. If an error is reported it will result in an HTTP
error code as expected. This does not change the behavior of errors thrown
AFTER data has been sent: They will still result in an HTTP 200 even if an
error is reported.
The line protocol between Erlang and os processes has been extended to support
an optional Header field on "chunks" and "end". The javascript list handling
has been updated to use this if a new header is set via start(...). This makes
it possible to begin processing with getRow(), but later reset the headers via
start(...). Again, if data has been sent(...) the new headers will NOT take
effect.
COUCHDB-430
COUCHDB-514
COUCHDB-764
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This patch breaks some update handlers and also assumes that any JSON
response body with a property called 'id' is the actual id of a
created or updated document.
This reverts;
0a64f310b43b90176bb59040e7e866978c429c88
4edbb93d2271ac1eb82f4d2bb072b8bdf6829f85
b9af7ea506ac95034a91951f29e1f4249ef4317c
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This lets the OS X startup define this user's config which will be
used for editing and what-not.
Forward port from 1.2.0
Change-Id: I03ebce57030457e25c9c8cb60a679547a0cd84bf
Reviewed-on: http://review.membase.org/6380
Tested-by: Dustin Sallings <dustin@spy.net>
Reviewed-by: Steve Yen <steve.yen@gmail.com>
Reviewed-by: Filipe David Borba Manana <fdmanana@gmail.com>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
The parts of a couchdb authentication cookie are separated by
colons. One of these parts can contain colons and, more rarely, runs
of colons. The string:tokens function silently drops any empty token,
thus giving a spurious failure for valid input. The fix changes this
mechanism to one that losslessly decodes this part.
COUCHDB-1607
|
|
|
|
|
|
|
|
|
|
|
|
| |
This can be used to serve up files like /crossdomain.xml from a
CouchDB instance. For example, you could add:
[httpd_global_handlers]
crossdomain.xml = {couch_httpd_misc_handlers, handle_file_req, "/usr/share/couchdb/www/crossdomain.xml" }
To your local.ini to serve up crossdomain.xml.
Closes #4
|
|
|
|
|
|
|
|
|
| |
This patch introduces a stable server-wide UUID which is used in place
of the local hostname and port number in new replication ids. This
allows CouchDB to find a valid checkpoint even if the coordinating
node's port has changed (it might be using a dynamic port, for example).
COUCHDB-1259
|
|
|
|
|
|
| |
Turns out that etap consumes any message in the mailbox in
some cases. This can make some tests that use message passing
hang, as etap itself consumes the messages.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
With this patch, you can set log levels per CouchDB module that
overrides the default set in `[log] level = `.
For example:
[log]
level = info
[log_level_by_module]
couch_httpd = debug
This will have all modules log at level 'info' and `couch_httpd` log
at level 'debug'.
See src/*/*.erl for the various CouchDB modules.
Based on work started by Robert Newson.
|
| |
|
|
|
|
| |
THIS NEEDS A PROPER FIX.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
The flow control in ibrowse's async response streams is tricky. We call
stream_next to pull more data off the socket, but it seems that ibrowse
will sometimes split that data into multiple messages. If we call
stream_next for each message we process we end up with an overflowing
mailbox.
This patch changes the consumer so that it clears out the mailbox before
calling stream_next.
|
| |
|
|
|
|
| |
...but log a message instead. Fixes COUCHDB-1445.
|
|
|
|
|
|
|
|
|
|
| |
This change should reduce the risk that administrators fail to
recognize correctable server errors without creating excess noise.
If malformed requests start throwing 5xx errors into the logs, we
should look at detecting them and reporting a better error, probably
a 4xx, but it's more important that an admin known that, e.g., the
permissions of a .couch file are wrong.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It's better to let these errors bubble and/or not give them special
treatment when file:format_error/1 can do a better job of describing
the failure.
This is a forward-port of work done on 1.2.x, encompassing the
following commits:
04c4a1041de06e8e39983cf9616c8baf9d889106
ede9482fc3c9e629572ab1376d88f4499c4c8beb
cd238b42d1333cece5ab899c35c3b83ada8d448a
af9e1c883ca4accb209cccc7ac9d26efa4daf1fd
ba8fa86890087650a02a8ebaaea2537468999472
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
the _dst variant does exist in R14B04.
|
|
|
|
|
|
| |
Acceptor cannot be recycled until ssl handshake is done,
so it's possible and easy for all acceptors be blocked between the point
where the new socket's connected and the ssl handshake is done.
|
| |
|
|
|
|
|
| |
Instead of replacing 'EXIT' signals we can just stuff the process that
exited into the exit reason and match on that.
|
|
|
|
| |
This reverts commit f5341a7dc65e96b806a958a27c9ccee5ec431162.
|