path: root/doc/src/sgml/logicaldecoding.sgml

<!-- doc/src/sgml/logicaldecoding.sgml -->
 <chapter id="logicaldecoding">
  <title>Logical Decoding</title>
  <indexterm zone="logicaldecoding">
   <primary>Logical Decoding</primary>
  </indexterm>
  <para>
   PostgreSQL provides infrastructure to stream the modifications performed
   via SQL to external consumers.  This functionality can be used for a
   variety of purposes, including replication solutions and auditing.
  </para>

  <para>
   Changes are sent out in streams identified by logical replication slots.
  </para>

  <para>
   The format in which those changes are streamed is determined by the output
   plugin used.  An example plugin is provided in the PostgreSQL distribution.
   Additional plugins can be
   written to extend the choice of available formats without modifying any
   core code.
   Every output plugin has access to each individual new row produced
   by <command>INSERT</command> and the new row version created
   by <command>UPDATE</command>.  Availability of old row versions for
   <command>UPDATE</command> and <command>DELETE</command> depends on
   the configured replica identity (see <xref linkend="sql-altertable-replica-identity"/>).
  </para>

  <para>
   Changes can be consumed either using the streaming replication protocol
   (see <xref linkend="protocol-replication"/> and
   <xref linkend="logicaldecoding-walsender"/>), or by calling functions
   via SQL (see <xref linkend="logicaldecoding-sql"/>). It is also possible
   to write additional methods of consuming the output of a replication slot
   without modifying core code
   (see <xref linkend="logicaldecoding-writer"/>).
  </para>

  <sect1 id="logicaldecoding-example">
   <title>Logical Decoding Examples</title>

   <para>
    The following example demonstrates controlling logical decoding using the
    SQL interface.
   </para>

   <para>
    Before you can use logical decoding, you must set
    <xref linkend="guc-wal-level"/> to <literal>logical</literal> and
    <xref linkend="guc-max-replication-slots"/> to at least 1.  Then, you
    should connect to the target database (in the example
    below, <literal>postgres</literal>) as a superuser.
   </para>

<programlisting>
postgres=# -- Create a slot named 'regression_slot' using the output plugin 'test_decoding'
postgres=# SELECT * FROM pg_create_logical_replication_slot('regression_slot', 'test_decoding', false, true);
    slot_name    |    lsn
-----------------+-----------
 regression_slot | 0/16B1970
(1 row)

postgres=# SELECT slot_name, plugin, slot_type, database, active, restart_lsn, confirmed_flush_lsn FROM pg_replication_slots;
    slot_name    |    plugin     | slot_type | database | active | restart_lsn | confirmed_flush_lsn
-----------------+---------------+-----------+----------+--------+-------------+-----------------
 regression_slot | test_decoding | logical   | postgres | f      | 0/16A4408   | 0/16A4440
(1 row)

postgres=# -- There are no changes to see yet
postgres=# SELECT * FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL);
 lsn | xid | data
-----+-----+------
(0 rows)

postgres=# CREATE TABLE data(id serial primary key, data text);
CREATE TABLE

postgres=# -- DDL isn't replicated, so all you'll see is the transaction
postgres=# SELECT * FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL);
    lsn    |  xid  |     data
-----------+-------+--------------
 0/BA2DA58 | 10297 | BEGIN 10297
 0/BA5A5A0 | 10297 | COMMIT 10297
(2 rows)

postgres=# -- Once changes are read, they're consumed and not emitted
postgres=# -- in a subsequent call:
postgres=# SELECT * FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL);
 lsn | xid | data
-----+-----+------
(0 rows)

postgres=# BEGIN;
postgres=*# INSERT INTO data(data) VALUES('1');
postgres=*# INSERT INTO data(data) VALUES('2');
postgres=*# COMMIT;

postgres=# SELECT * FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL);
    lsn    |  xid  |                          data
-----------+-------+---------------------------------------------------------
 0/BA5A688 | 10298 | BEGIN 10298
 0/BA5A6F0 | 10298 | table public.data: INSERT: id[integer]:1 data[text]:'1'
 0/BA5A7F8 | 10298 | table public.data: INSERT: id[integer]:2 data[text]:'2'
 0/BA5A8A8 | 10298 | COMMIT 10298
(4 rows)

postgres=# INSERT INTO data(data) VALUES('3');

postgres=# -- You can also peek ahead in the change stream without consuming changes
postgres=# SELECT * FROM pg_logical_slot_peek_changes('regression_slot', NULL, NULL);
    lsn    |  xid  |                          data
-----------+-------+---------------------------------------------------------
 0/BA5A8E0 | 10299 | BEGIN 10299
 0/BA5A8E0 | 10299 | table public.data: INSERT: id[integer]:3 data[text]:'3'
 0/BA5A990 | 10299 | COMMIT 10299
(3 rows)

postgres=# -- The next call to pg_logical_slot_peek_changes() returns the same changes again
postgres=# SELECT * FROM pg_logical_slot_peek_changes('regression_slot', NULL, NULL);
    lsn    |  xid  |                          data
-----------+-------+---------------------------------------------------------
 0/BA5A8E0 | 10299 | BEGIN 10299
 0/BA5A8E0 | 10299 | table public.data: INSERT: id[integer]:3 data[text]:'3'
 0/BA5A990 | 10299 | COMMIT 10299
(3 rows)

postgres=# -- options can be passed to output plugin, to influence the formatting
postgres=# SELECT * FROM pg_logical_slot_peek_changes('regression_slot', NULL, NULL, 'include-timestamp', 'on');
    lsn    |  xid  |                          data
-----------+-------+---------------------------------------------------------
 0/BA5A8E0 | 10299 | BEGIN 10299
 0/BA5A8E0 | 10299 | table public.data: INSERT: id[integer]:3 data[text]:'3'
 0/BA5A990 | 10299 | COMMIT 10299 (at 2017-05-10 12:07:21.272494-04)
(3 rows)

postgres=# -- Remember to destroy a slot you no longer need to stop it consuming
postgres=# -- server resources:
postgres=# SELECT pg_drop_replication_slot('regression_slot');
 pg_drop_replication_slot
-----------------------

(1 row)
</programlisting>

   <para>
    The following examples shows how logical decoding is controlled over the
    streaming replication protocol, using the
    program <xref linkend="app-pgrecvlogical"/> included in the PostgreSQL
    distribution.  This requires that client authentication is set up to allow
    replication connections
    (see <xref linkend="streaming-replication-authentication"/>) and
    that <varname>max_wal_senders</varname> is set sufficiently high to allow
    an additional connection.  The second example shows how to stream two-phase
    transactions.  Before you use two-phase commands, you must set
    <xref linkend="guc-max-prepared-transactions"/> to at least 1.
   </para>
<programlisting>
Example 1:
$ pg_recvlogical -d postgres --slot=test --create-slot
$ pg_recvlogical -d postgres --slot=test --start -f -
<keycombo action="simul"><keycap>Control</keycap><keycap>Z</keycap></keycombo>
$ psql -d postgres -c "INSERT INTO data(data) VALUES('4');"
$ fg
BEGIN 693
table public.data: INSERT: id[integer]:4 data[text]:'4'
COMMIT 693
<keycombo action="simul"><keycap>Control</keycap><keycap>C</keycap></keycombo>
$ pg_recvlogical -d postgres --slot=test --drop-slot

Example 2:
$ pg_recvlogical -d postgres --slot=test --create-slot --two-phase
$ pg_recvlogical -d postgres --slot=test --start -f -
<keycombo action="simul"><keycap>Control</keycap><keycap>Z</keycap></keycombo>
$ psql -d postgres -c "BEGIN;INSERT INTO data(data) VALUES('5');PREPARE TRANSACTION 'test';"
$ fg
BEGIN 694
table public.data: INSERT: id[integer]:5 data[text]:'5'
PREPARE TRANSACTION 'test', txid 694
<keycombo action="simul"><keycap>Control</keycap><keycap>Z</keycap></keycombo>
$ psql -d postgres -c "COMMIT PREPARED 'test';"
$ fg
COMMIT PREPARED 'test', txid 694
<keycombo action="simul"><keycap>Control</keycap><keycap>C</keycap></keycombo>
$ pg_recvlogical -d postgres --slot=test --drop-slot
</programlisting>

  <para>
  The following example shows SQL interface that can be used to decode prepared
  transactions. Before you use two-phase commit commands, you must set
  <varname>max_prepared_transactions</varname> to at least 1. You must also have
  set the two-phase parameter as 'true' while creating the slot using
  <function>pg_create_logical_replication_slot</function>
  Note that we will stream the entire transaction after the commit if it
  is not already decoded.
  </para>
<programlisting>
postgres=# BEGIN;
postgres=*# INSERT INTO data(data) VALUES('5');
postgres=*# PREPARE TRANSACTION 'test_prepared1';

postgres=# SELECT * FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL);
    lsn    | xid |                          data
-----------+-----+---------------------------------------------------------
 0/1689DC0 | 529 | BEGIN 529
 0/1689DC0 | 529 | table public.data: INSERT: id[integer]:3 data[text]:'5'
 0/1689FC0 | 529 | PREPARE TRANSACTION 'test_prepared1', txid 529
(3 rows)

postgres=# COMMIT PREPARED 'test_prepared1';
postgres=# select * from pg_logical_slot_get_changes('regression_slot', NULL, NULL);
    lsn    | xid |                    data
-----------+-----+--------------------------------------------
 0/168A060 | 529 | COMMIT PREPARED 'test_prepared1', txid 529
(4 row)

postgres=#-- you can also rollback a prepared transaction
postgres=# BEGIN;
postgres=*# INSERT INTO data(data) VALUES('6');
postgres=*# PREPARE TRANSACTION 'test_prepared2';
postgres=# select * from pg_logical_slot_get_changes('regression_slot', NULL, NULL);
    lsn    | xid |                          data
-----------+-----+---------------------------------------------------------
 0/168A180 | 530 | BEGIN 530
 0/168A1E8 | 530 | table public.data: INSERT: id[integer]:4 data[text]:'6'
 0/168A430 | 530 | PREPARE TRANSACTION 'test_prepared2', txid 530
(3 rows)

postgres=# ROLLBACK PREPARED 'test_prepared2';
postgres=# select * from pg_logical_slot_get_changes('regression_slot', NULL, NULL);
    lsn    | xid |                     data
-----------+-----+----------------------------------------------
 0/168A4B8 | 530 | ROLLBACK PREPARED 'test_prepared2', txid 530
(1 row)
</programlisting>
</sect1>

  <sect1 id="logicaldecoding-explanation">
   <title>Logical Decoding Concepts</title>
   <sect2 id="logicaldecoding-explanation-log-dec">
    <title>Logical Decoding</title>

    <indexterm>
     <primary>Logical Decoding</primary>
    </indexterm>

    <para>
     Logical decoding is the process of extracting all persistent changes
     to a database's tables into a coherent, easy to understand format which
     can be interpreted without detailed knowledge of the database's internal
     state.
    </para>

    <para>
     In <productname>PostgreSQL</productname>, logical decoding is implemented
     by decoding the contents of the <link linkend="wal">write-ahead
     log</link>, which describe changes on a storage level, into an
     application-specific form such as a stream of tuples or SQL statements.
    </para>
   </sect2>

   <sect2 id="logicaldecoding-replication-slots">
    <title>Replication Slots</title>

    <indexterm>
     <primary>replication slot</primary>
     <secondary>logical replication</secondary>
    </indexterm>

    <para>
     In the context of logical replication, a slot represents a stream of
     changes that can be replayed to a client in the order they were made on
     the origin server. Each slot streams a sequence of changes from a single
     database.
    </para>

    <note>
     <para><productname>PostgreSQL</productname> also has streaming replication slots
     (see <xref linkend="streaming-replication"/>), but they are used somewhat
     differently there.
     </para>
    </note>

    <para>
     A replication slot has an identifier that is unique across all databases
     in a <productname>PostgreSQL</productname> cluster. Slots persist
     independently of the connection using them and are crash-safe.
    </para>

    <para>
     A logical slot will emit each change just once in normal operation.
     The current position of each slot is persisted only at checkpoint, so in
     the case of a crash the slot may return to an earlier LSN, which will
     then cause recent changes to be sent again when the server restarts.
     Logical decoding clients are responsible for avoiding ill effects from
     handling the same message more than once.  Clients may wish to record
     the last LSN they saw when decoding and skip over any repeated data or
     (when using the replication protocol) request that decoding start from
     that LSN rather than letting the server determine the start point.
     The Replication Progress Tracking feature is designed for this purpose,
     refer to <link linkend="replication-origins">replication origins</link>.
    </para>

    <para>
     Multiple independent slots may exist for a single database. Each slot has
     its own state, allowing different consumers to receive changes from
     different points in the database change stream. For most applications, a
     separate slot will be required for each consumer.
    </para>

    <para>
     A logical replication slot knows nothing about the state of the
     receiver(s).  It's even possible to have multiple different receivers using
     the same slot at different times; they'll just get the changes following
     on from when the last receiver stopped consuming them. Only one receiver
     may consume changes from a slot at any given time.
    </para>

    <para>
     A logical replication slot can also be created on a hot standby. To prevent
     <command>VACUUM</command> from removing required rows from the system
     catalogs, <varname>hot_standby_feedback</varname> should be set on the
     standby. In spite of that, if any required rows get removed, the slot gets
     invalidated. It's highly recommended to use a physical slot between the
     primary and the standby. Otherwise, <varname>hot_standby_feedback</varname>
     will work but only while the connection is alive (for example a node
     restart would break it). Then, the primary may delete system catalog rows
     that could be needed by the logical decoding on the standby (as it does
     not know about the catalog_xmin on the standby). Existing logical slots
     on standby also get invalidated if <varname>wal_level</varname> on the
     primary is reduced to less than <literal>logical</literal>.
     This is done as soon as the standby detects such a change in the WAL stream.
     It means that, for walsenders which are lagging (if any), some WAL records up
     to the <varname>wal_level</varname> parameter change on the primary won't be
     decoded.
    </para>

    <para>
     Creation of a logical slot requires information about all the currently
     running transactions. On the primary, this information is available
     directly, but on a standby, this information has to be obtained from
     primary. Thus, slot creation may need to wait for some activity to happen
     on the primary. If the primary is idle, creating a logical slot on
     standby may take noticeable time. This can be sped up by calling the
     <function>pg_log_standby_snapshot</function> function on the primary.
    </para>

    <caution>
     <para>
      Replication slots persist across crashes and know nothing about the state
      of their consumer(s). They will prevent removal of required resources
      even when there is no connection using them. This consumes storage
      because neither required WAL nor required rows from the system catalogs
      can be removed by <command>VACUUM</command> as long as they are required by a replication
      slot.  In extreme cases this could cause the database to shut down to prevent
      transaction ID wraparound (see <xref linkend="vacuum-for-wraparound"/>).
      So if a slot is no longer required it should be dropped.
     </para>
    </caution>
   </sect2>

   <sect2 id="logicaldecoding-explanation-output-plugins">
    <title>Output Plugins</title>
    <para>
     Output plugins transform the data from the write-ahead log's internal
     representation into the format the consumer of a replication slot desires.
    </para>
   </sect2>

   <sect2 id="logicaldecoding-explanation-exported-snapshots">
    <title>Exported Snapshots</title>
    <para>
     When a new replication slot is created using the streaming replication
     interface (see <xref linkend="protocol-replication-create-replication-slot"/>), a
     snapshot is exported
     (see <xref linkend="functions-snapshot-synchronization"/>), which will show
     exactly the state of the database after which all changes will be
     included in the change stream. This can be used to create a new replica by
     using <link linkend="sql-set-transaction"><literal>SET TRANSACTION
     SNAPSHOT</literal></link> to read the state of the database at the moment
     the slot was created. This transaction can then be used to dump the
     database's state at that point in time, which afterwards can be updated
     using the slot's contents without losing any changes.
    </para>
    <para>
     Creation of a snapshot is not always possible.  In particular, it will
     fail when connected to a hot standby.  Applications that do not require
     snapshot export may suppress it with the <literal>NOEXPORT_SNAPSHOT</literal>
     option.
    </para>
   </sect2>
  </sect1>

  <sect1 id="logicaldecoding-walsender">
   <title>Streaming Replication Protocol Interface</title>

   <para>
    The commands
    <itemizedlist>
     <listitem>
      <para><literal>CREATE_REPLICATION_SLOT <replaceable>slot_name</replaceable> LOGICAL <replaceable>output_plugin</replaceable></literal></para>
     </listitem>

     <listitem>
      <para><literal>DROP_REPLICATION_SLOT <replaceable>slot_name</replaceable></literal> <optional> <literal>WAIT</literal> </optional></para>
     </listitem>

     <listitem>
      <para><literal>START_REPLICATION SLOT <replaceable>slot_name</replaceable> LOGICAL ...</literal></para>
     </listitem>
    </itemizedlist>
    are used to create, drop, and stream changes from a replication
    slot, respectively. These commands are only available over a replication
    connection; they cannot be used via SQL.
    See <xref linkend="protocol-replication"/> for details on these commands.
   </para>

   <para>
    The command <xref linkend="app-pgrecvlogical"/> can be used to control
    logical decoding over a streaming replication connection.  (It uses
    these commands internally.)
   </para>
  </sect1>

  <sect1 id="logicaldecoding-sql">
   <title>Logical Decoding <acronym>SQL</acronym> Interface</title>

   <para>
     See <xref linkend="functions-replication"/> for detailed documentation on
     the SQL-level API for interacting with logical decoding.
   </para>

   <para>
    Synchronous replication (see <xref linkend="synchronous-replication"/>) is
    only supported on replication slots used over the streaming replication interface. The
    function interface and additional, non-core interfaces do not support
    synchronous replication.
   </para>
  </sect1>

  <sect1 id="logicaldecoding-catalogs">
   <title>System Catalogs Related to Logical Decoding</title>

   <para>
    The <link linkend="view-pg-replication-slots"><structname>pg_replication_slots</structname></link>
    view and the
    <link linkend="monitoring-pg-stat-replication-view">
    <structname>pg_stat_replication</structname></link>
    view provide information about the current state of replication slots and
    streaming replication connections respectively. These views apply to both physical and
    logical replication. The
    <link linkend="monitoring-pg-stat-replication-slots-view">
    <structname>pg_stat_replication_slots</structname></link>
    view provides statistics information about the logical replication slots.
   </para>
  </sect1>

  <sect1 id="logicaldecoding-output-plugin">
   <title>Logical Decoding Output Plugins</title>
   <para>
    An example output plugin can be found in the
    <link linkend="test-decoding">
     <filename>contrib/test_decoding</filename>
    </link>
    subdirectory of the PostgreSQL source tree.
   </para>
   <sect2 id="logicaldecoding-output-init">
    <title>Initialization Function</title>
    <indexterm zone="logicaldecoding-output-init">
     <primary>_PG_output_plugin_init</primary>
    </indexterm>
    <para>
     An output plugin is loaded by dynamically loading a shared library with
     the output plugin's name as the library base name. The normal library
     search path is used to locate the library. To provide the required output
     plugin callbacks and to indicate that the library is actually an output
     plugin it needs to provide a function named
     <function>_PG_output_plugin_init</function>. This function is passed a
     struct that needs to be filled with the callback function pointers for
     individual actions.
<programlisting>
typedef struct OutputPluginCallbacks
{
    LogicalDecodeStartupCB startup_cb;
    LogicalDecodeBeginCB begin_cb;
    LogicalDecodeChangeCB change_cb;
    LogicalDecodeTruncateCB truncate_cb;
    LogicalDecodeCommitCB commit_cb;
    LogicalDecodeMessageCB message_cb;
    LogicalDecodeFilterByOriginCB filter_by_origin_cb;
    LogicalDecodeShutdownCB shutdown_cb;
    LogicalDecodeFilterPrepareCB filter_prepare_cb;
    LogicalDecodeBeginPrepareCB begin_prepare_cb;
    LogicalDecodePrepareCB prepare_cb;
    LogicalDecodeCommitPreparedCB commit_prepared_cb;
    LogicalDecodeRollbackPreparedCB rollback_prepared_cb;
    LogicalDecodeStreamStartCB stream_start_cb;
    LogicalDecodeStreamStopCB stream_stop_cb;
    LogicalDecodeStreamAbortCB stream_abort_cb;
    LogicalDecodeStreamPrepareCB stream_prepare_cb;
    LogicalDecodeStreamCommitCB stream_commit_cb;
    LogicalDecodeStreamChangeCB stream_change_cb;
    LogicalDecodeStreamMessageCB stream_message_cb;
    LogicalDecodeStreamTruncateCB stream_truncate_cb;
} OutputPluginCallbacks;

typedef void (*LogicalOutputPluginInit) (struct OutputPluginCallbacks *cb);
</programlisting>
     The <function>begin_cb</function>, <function>change_cb</function>
     and <function>commit_cb</function> callbacks are required,
     while <function>startup_cb</function>, <function>truncate_cb</function>,
     <function>message_cb</function>, <function>filter_by_origin_cb</function>,
     and <function>shutdown_cb</function> are optional.
     If <function>truncate_cb</function> is not set but a
     <command>TRUNCATE</command> is to be decoded, the action will be ignored.
    </para>

    <para>
     An output plugin may also define functions to support streaming of large,
     in-progress transactions. The <function>stream_start_cb</function>,
     <function>stream_stop_cb</function>, <function>stream_abort_cb</function>,
     <function>stream_commit_cb</function>, and <function>stream_change_cb</function>
     are required, while <function>stream_message_cb</function> and
     <function>stream_truncate_cb</function> are optional. The
     <function>stream_prepare_cb</function> is also required if the output
     plugin also support two-phase commits.
    </para>

    <para>
    An output plugin may also define functions to support two-phase commits,
    which allows actions to be decoded on the <command>PREPARE TRANSACTION</command>.
    The <function>begin_prepare_cb</function>, <function>prepare_cb</function>,
    <function>commit_prepared_cb</function> and <function>rollback_prepared_cb</function>
    callbacks are required, while <function>filter_prepare_cb</function> is optional.
    The <function>stream_prepare_cb</function> is also required if the output plugin
    also supports the streaming of large in-progress transactions.
    </para>
   </sect2>

   <sect2 id="logicaldecoding-capabilities">
    <title>Capabilities</title>

    <para>
     To decode, format and output changes, output plugins can use most of the
     backend's normal infrastructure, including calling output functions. Read
     only access to relations is permitted as long as only relations are
     accessed that either have been created by <command>initdb</command> in
     the <literal>pg_catalog</literal> schema, or have been marked as user
     provided catalog tables using
<programlisting>
ALTER TABLE user_catalog_table SET (user_catalog_table = true);
CREATE TABLE another_catalog_table(data text) WITH (user_catalog_table = true);
</programlisting>
     Note that access to user catalog tables or regular system catalog tables
     in the output plugins has to be done via the <literal>systable_*</literal>
     scan APIs only. Access via the <literal>heap_*</literal> scan APIs will
     error out. Additionally, any actions leading to transaction ID assignment
     are prohibited. That, among others, includes writing to tables, performing
     DDL changes, and calling <literal>pg_current_xact_id()</literal>.
    </para>
   </sect2>

   <sect2 id="logicaldecoding-output-mode">
    <title>Output Modes</title>

    <para>
     Output plugin callbacks can pass data to the consumer in nearly arbitrary
     formats. For some use cases, like viewing the changes via SQL, returning
     data in a data type that can contain arbitrary data (e.g., <type>bytea</type>) is
     cumbersome. If the output plugin only outputs textual data in the
     server's encoding, it can declare that by
     setting <literal>OutputPluginOptions.output_type</literal>
     to <literal>OUTPUT_PLUGIN_TEXTUAL_OUTPUT</literal> instead
     of <literal>OUTPUT_PLUGIN_BINARY_OUTPUT</literal> in
     the <link linkend="logicaldecoding-output-plugin-startup">startup
     callback</link>. In that case, all the data has to be in the server's encoding
     so that a <type>text</type> datum can contain it. This is checked in assertion-enabled
     builds.
    </para>
   </sect2>

   <sect2 id="logicaldecoding-output-plugin-callbacks">
    <title>Output Plugin Callbacks</title>

    <para>
     An output plugin gets notified about changes that are happening via
     various callbacks it needs to provide.
    </para>

    <para>
     Concurrent transactions are decoded in commit order, and only changes
     belonging to a specific transaction are decoded between
     the <literal>begin</literal> and <literal>commit</literal>
     callbacks. Transactions that were rolled back explicitly or implicitly
     never get
     decoded. Successful savepoints are
     folded into the transaction containing them in the order they were
     executed within that transaction. A transaction that is prepared for
     a two-phase commit using <command>PREPARE TRANSACTION</command> will
     also be decoded if the output plugin callbacks needed for decoding
     them are provided. It is possible that the current prepared transaction
     which is being decoded is aborted concurrently via a
     <command>ROLLBACK PREPARED</command> command. In that case, the logical
     decoding of this transaction will be aborted too. All the changes of such
     a transaction are skipped once the abort is detected and the
     <function>prepare_cb</function> callback is invoked. Thus even in case of
     a concurrent abort, enough information is provided to the output plugin
     for it to properly deal with <command>ROLLBACK PREPARED</command> once
     that is decoded.
    </para>

    <note>
     <para>
      Only transactions that have already safely been flushed to disk will be
      decoded. That can lead to a <command>COMMIT</command> not immediately being decoded in a
      directly following <literal>pg_logical_slot_get_changes()</literal>
      when <varname>synchronous_commit</varname> is set
      to <literal>off</literal>.
     </para>
    </note>

    <sect3 id="logicaldecoding-output-plugin-startup">
     <title>Startup Callback</title>
     <para>
      The optional <function>startup_cb</function> callback is called whenever
      a replication slot is created or asked to stream changes, independent
      of the number of changes that are ready to be put out.
<programlisting>
typedef void (*LogicalDecodeStartupCB) (struct LogicalDecodingContext *ctx,
                                        OutputPluginOptions *options,
                                        bool is_init);
</programlisting>
      The <literal>is_init</literal> parameter will be true when the
      replication slot is being created and false
      otherwise. <parameter>options</parameter> points to a struct of options
      that output plugins can set:
<programlisting>
typedef struct OutputPluginOptions
{
    OutputPluginOutputType output_type;
    bool        receive_rewrites;
} OutputPluginOptions;
</programlisting>
      <literal>output_type</literal> has to either be set to
      <literal>OUTPUT_PLUGIN_TEXTUAL_OUTPUT</literal>
      or <literal>OUTPUT_PLUGIN_BINARY_OUTPUT</literal>. See also
      <xref linkend="logicaldecoding-output-mode"/>.
      If <literal>receive_rewrites</literal> is true, the output plugin will
      also be called for changes made by heap rewrites during certain DDL
      operations.  These are of interest to plugins that handle DDL
      replication, but they require special handling.
     </para>

     <para>
      The startup callback should validate the options present in
      <literal>ctx-&gt;output_plugin_options</literal>. If the output plugin
      needs to have a state, it can
      use <literal>ctx-&gt;output_plugin_private</literal> to store it.
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-shutdown">
     <title>Shutdown Callback</title>

     <para>
      The optional <function>shutdown_cb</function> callback is called
      whenever a formerly active replication slot is not used anymore and can
      be used to deallocate resources private to the output plugin. The slot
      isn't necessarily being dropped, streaming is just being stopped.
<programlisting>
typedef void (*LogicalDecodeShutdownCB) (struct LogicalDecodingContext *ctx);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-begin">
     <title>Transaction Begin Callback</title>

     <para>
      The required <function>begin_cb</function> callback is called whenever a
      start of a committed transaction has been decoded. Aborted transactions
      and their contents never get decoded.
<programlisting>
typedef void (*LogicalDecodeBeginCB) (struct LogicalDecodingContext *ctx,
                                      ReorderBufferTXN *txn);
</programlisting>
      The <parameter>txn</parameter> parameter contains meta information about
      the transaction, like the time stamp at which it has been committed and
      its XID.
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-commit">
     <title>Transaction End Callback</title>

     <para>
      The required <function>commit_cb</function> callback is called whenever
      a transaction commit has been
      decoded. The <function>change_cb</function> callbacks for all modified
      rows will have been called before this, if there have been any modified
      rows.
<programlisting>
typedef void (*LogicalDecodeCommitCB) (struct LogicalDecodingContext *ctx,
                                       ReorderBufferTXN *txn,
                                       XLogRecPtr commit_lsn);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-change">
     <title>Change Callback</title>

     <para>
      The required <function>change_cb</function> callback is called for every
      individual row modification inside a transaction, may it be
      an <command>INSERT</command>, <command>UPDATE</command>,
      or <command>DELETE</command>. Even if the original command modified
      several rows at once the callback will be called individually for each
      row. The <function>change_cb</function> callback may access system or
      user catalog tables to aid in the process of outputting the row
      modification details. In case of decoding a prepared (but yet
      uncommitted) transaction or decoding of an uncommitted transaction, this
      change callback might also error out due to simultaneous rollback of
      this very same transaction. In that case, the logical decoding of this
      aborted transaction is stopped gracefully.
<programlisting>
typedef void (*LogicalDecodeChangeCB) (struct LogicalDecodingContext *ctx,
                                       ReorderBufferTXN *txn,
                                       Relation relation,
                                       ReorderBufferChange *change);
</programlisting>
      The <parameter>ctx</parameter> and <parameter>txn</parameter> parameters
      have the same contents as for the <function>begin_cb</function>
      and <function>commit_cb</function> callbacks, but additionally the
      relation descriptor <parameter>relation</parameter> points to the
      relation the row belongs to and a struct
      <parameter>change</parameter> describing the row modification are passed
      in.
     </para>

     <note>
      <para>
       Only changes in user defined tables that are not unlogged
       (see <xref linkend="sql-createtable-unlogged"/>) and not temporary
       (see <xref linkend="sql-createtable-temporary"/>) can be extracted using
       logical decoding.
      </para>
     </note>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-truncate">
     <title>Truncate Callback</title>

     <para>
      The optional <function>truncate_cb</function> callback is called for a
      <command>TRUNCATE</command> command.
<programlisting>
typedef void (*LogicalDecodeTruncateCB) (struct LogicalDecodingContext *ctx,
                                         ReorderBufferTXN *txn,
                                         int nrelations,
                                         Relation relations[],
                                         ReorderBufferChange *change);
</programlisting>
      The parameters are analogous to the <function>change_cb</function>
      callback.  However, because <command>TRUNCATE</command> actions on
      tables connected by foreign keys need to be executed together, this
      callback receives an array of relations instead of just a single one.
      See the description of the <xref linkend="sql-truncate"/> statement for
      details.
     </para>
    </sect3>

     <sect3 id="logicaldecoding-output-plugin-filter-origin">
     <title>Origin Filter Callback</title>

     <para>
       The optional <function>filter_by_origin_cb</function> callback
       is called to determine whether data that has been replayed
       from <parameter>origin_id</parameter> is of interest to the
       output plugin.
<programlisting>
typedef bool (*LogicalDecodeFilterByOriginCB) (struct LogicalDecodingContext *ctx,
                                               RepOriginId origin_id);
</programlisting>
      The <parameter>ctx</parameter> parameter has the same contents
      as for the other callbacks. No information but the origin is
      available. To signal that changes originating on the passed in
      node are irrelevant, return true, causing them to be filtered
      away; false otherwise. The other callbacks will not be called
      for transactions and changes that have been filtered away.
     </para>
     <para>
       This is useful when implementing cascading or multidirectional
       replication solutions. Filtering by the origin allows to
       prevent replicating the same changes back and forth in such
       setups.  While transactions and changes also carry information
       about the origin, filtering via this callback is noticeably
       more efficient.
     </para>
     </sect3>

    <sect3 id="logicaldecoding-output-plugin-message">
     <title>Generic Message Callback</title>

     <para>
      The optional <function>message_cb</function> callback is called whenever
      a logical decoding message has been decoded.
<programlisting>
typedef void (*LogicalDecodeMessageCB) (struct LogicalDecodingContext *ctx,
                                        ReorderBufferTXN *txn,
                                        XLogRecPtr message_lsn,
                                        bool transactional,
                                        const char *prefix,
                                        Size message_size,
                                        const char *message);
</programlisting>
      The <parameter>txn</parameter> parameter contains meta information about
      the transaction, like the time stamp at which it has been committed and
      its XID. Note however that it can be NULL when the message is
      non-transactional and the XID was not assigned yet in the transaction
      which logged the message. The <parameter>lsn</parameter> has WAL
      location of the message. The <parameter>transactional</parameter> says
      if the message was sent as transactional or not. Similar to the change
      callback, in case of decoding a prepared (but yet uncommitted)
      transaction or decoding of an uncommitted transaction, this message
      callback might also error out due to simultaneous rollback of
      this very same transaction. In that case, the logical decoding of this
      aborted transaction is stopped gracefully.

      The <parameter>prefix</parameter> is arbitrary null-terminated prefix
      which can be used for identifying interesting messages for the current
      plugin. And finally the <parameter>message</parameter> parameter holds
      the actual message of <parameter>message_size</parameter> size.
     </para>
     <para>
      Extra care should be taken to ensure that the prefix the output plugin
      considers interesting is unique. Using name of the extension or the
      output plugin itself is often a good choice.
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-filter-prepare">
     <title>Prepare Filter Callback</title>

     <para>
       The optional <function>filter_prepare_cb</function> callback
       is called to determine whether data that is part of the current
       two-phase commit transaction should be considered for decoding
       at this prepare stage or later as a regular one-phase transaction at
       <command>COMMIT PREPARED</command> time. To signal that
       decoding should be skipped, return <literal>true</literal>;
       <literal>false</literal> otherwise. When the callback is not
       defined, <literal>false</literal> is assumed (i.e. no filtering, all
       transactions using two-phase commit are decoded in two phases as well).
<programlisting>
typedef bool (*LogicalDecodeFilterPrepareCB) (struct LogicalDecodingContext *ctx,
                                              TransactionId xid,
                                              const char *gid);
</programlisting>
       The <parameter>ctx</parameter> parameter has the same contents as for
       the other callbacks. The parameters <parameter>xid</parameter>
       and <parameter>gid</parameter> provide two different ways to identify
       the transaction.  The later <command>COMMIT PREPARED</command> or
       <command>ROLLBACK PREPARED</command> carries both identifiers,
       providing an output plugin the choice of what to use.
     </para>
     <para>
       The callback may be invoked multiple times per transaction to decode
       and must provide the same static answer for a given pair of
       <parameter>xid</parameter> and <parameter>gid</parameter> every time
       it is called.
     </para>
     </sect3>

    <sect3 id="logicaldecoding-output-plugin-begin-prepare">
     <title>Transaction Begin Prepare Callback</title>

     <para>
      The required <function>begin_prepare_cb</function> callback is called
      whenever the start of a prepared transaction has been decoded. The
      <parameter>gid</parameter> field, which is part of the
      <parameter>txn</parameter> parameter, can be used in this callback to
      check if the plugin has already received this <command>PREPARE</command>
      in which case it can either error out or skip the remaining changes of
      the transaction.
<programlisting>
typedef void (*LogicalDecodeBeginPrepareCB) (struct LogicalDecodingContext *ctx,
                                             ReorderBufferTXN *txn);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-prepare">
     <title>Transaction Prepare Callback</title>

     <para>
      The required <function>prepare_cb</function> callback is called whenever
      a transaction which is prepared for two-phase commit has been
      decoded. The <function>change_cb</function> callback for all modified
      rows will have been called before this, if there have been any modified
      rows. The <parameter>gid</parameter> field, which is part of the
      <parameter>txn</parameter> parameter, can be used in this callback.
<programlisting>
typedef void (*LogicalDecodePrepareCB) (struct LogicalDecodingContext *ctx,
                                        ReorderBufferTXN *txn,
                                        XLogRecPtr prepare_lsn);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-commit-prepared">
     <title>Transaction Commit Prepared Callback</title>

     <para>
      The required <function>commit_prepared_cb</function> callback is called
      whenever a transaction <command>COMMIT PREPARED</command> has been decoded.
      The <parameter>gid</parameter> field, which is part of the
      <parameter>txn</parameter> parameter, can be used in this callback.
<programlisting>
typedef void (*LogicalDecodeCommitPreparedCB) (struct LogicalDecodingContext *ctx,
                                               ReorderBufferTXN *txn,
                                               XLogRecPtr commit_lsn);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-rollback-prepared">
     <title>Transaction Rollback Prepared Callback</title>

     <para>
      The required <function>rollback_prepared_cb</function> callback is called
      whenever a transaction <command>ROLLBACK PREPARED</command> has been
      decoded. The <parameter>gid</parameter> field, which is part of the
      <parameter>txn</parameter> parameter, can be used in this callback. The
      parameters <parameter>prepare_end_lsn</parameter> and
      <parameter>prepare_time</parameter> can be used to check if the plugin
      has received this <command>PREPARE TRANSACTION</command> in which case
      it can apply the rollback, otherwise, it can skip the rollback operation. The
      <parameter>gid</parameter> alone is not sufficient because the downstream
      node can have a prepared transaction with same identifier.
<programlisting>
typedef void (*LogicalDecodeRollbackPreparedCB) (struct LogicalDecodingContext *ctx,
                                                 ReorderBufferTXN *txn,
                                                 XLogRecPtr prepare_end_lsn,
                                                 TimestampTz prepare_time);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-stream-start">
     <title>Stream Start Callback</title>
     <para>
      The required <function>stream_start_cb</function> callback is called when
      opening a block of streamed changes from an in-progress transaction.
<programlisting>
typedef void (*LogicalDecodeStreamStartCB) (struct LogicalDecodingContext *ctx,
                                            ReorderBufferTXN *txn);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-stream-stop">
     <title>Stream Stop Callback</title>
     <para>
      The required <function>stream_stop_cb</function> callback is called when
      closing a block of streamed changes from an in-progress transaction.
<programlisting>
typedef void (*LogicalDecodeStreamStopCB) (struct LogicalDecodingContext *ctx,
                                           ReorderBufferTXN *txn);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-stream-abort">
     <title>Stream Abort Callback</title>
     <para>
      The required <function>stream_abort_cb</function> callback is called to
      abort a previously streamed transaction.
<programlisting>
typedef void (*LogicalDecodeStreamAbortCB) (struct LogicalDecodingContext *ctx,
                                            ReorderBufferTXN *txn,
                                            XLogRecPtr abort_lsn);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-stream-prepare">
     <title>Stream Prepare Callback</title>
     <para>
      The <function>stream_prepare_cb</function> callback is called to prepare
      a previously streamed transaction as part of a two-phase commit. This
      callback is required when the output plugin supports both the streaming
      of large in-progress transactions and two-phase commits.
      <programlisting>
typedef void (*LogicalDecodeStreamPrepareCB) (struct LogicalDecodingContext *ctx,
                                              ReorderBufferTXN *txn,
                                              XLogRecPtr prepare_lsn);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-stream-commit">
     <title>Stream Commit Callback</title>
     <para>
      The required <function>stream_commit_cb</function> callback is called to
      commit a previously streamed transaction.
<programlisting>
typedef void (*LogicalDecodeStreamCommitCB) (struct LogicalDecodingContext *ctx,
                                             ReorderBufferTXN *txn,
                                             XLogRecPtr commit_lsn);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-stream-change">
     <title>Stream Change Callback</title>
     <para>
      The required <function>stream_change_cb</function> callback is called
      when sending a change in a block of streamed changes (demarcated by
      <function>stream_start_cb</function> and <function>stream_stop_cb</function> calls).
      The actual changes are not displayed as the transaction can abort at a later
      point in time and we don't decode changes for aborted transactions.
<programlisting>
typedef void (*LogicalDecodeStreamChangeCB) (struct LogicalDecodingContext *ctx,
                                             ReorderBufferTXN *txn,
                                             Relation relation,
                                             ReorderBufferChange *change);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-stream-message">
     <title>Stream Message Callback</title>
     <para>
      The optional <function>stream_message_cb</function> callback is called when
      sending a generic message in a block of streamed changes (demarcated by
      <function>stream_start_cb</function> and <function>stream_stop_cb</function> calls).
      The message contents for transactional messages are not displayed as the transaction
      can abort at a later point in time and we don't decode changes for aborted
      transactions.
<programlisting>
typedef void (*LogicalDecodeStreamMessageCB) (struct LogicalDecodingContext *ctx,
                                              ReorderBufferTXN *txn,
                                              XLogRecPtr message_lsn,
                                              bool transactional,
                                              const char *prefix,
                                              Size message_size,
                                              const char *message);
</programlisting>
     </para>
    </sect3>

    <sect3 id="logicaldecoding-output-plugin-stream-truncate">
     <title>Stream Truncate Callback</title>
     <para>
      The optional <function>stream_truncate_cb</function> callback is called
      for a <command>TRUNCATE</command> command in a block of streamed changes
      (demarcated by <function>stream_start_cb</function> and
      <function>stream_stop_cb</function> calls).
<programlisting>
typedef void (*LogicalDecodeStreamTruncateCB) (struct LogicalDecodingContext *ctx,
                                               ReorderBufferTXN *txn,
                                               int nrelations,
                                               Relation relations[],
                                               ReorderBufferChange *change);
</programlisting>
      The parameters are analogous to the <function>stream_change_cb</function>
      callback.  However, because <command>TRUNCATE</command> actions on
      tables connected by foreign keys need to be executed together, this
      callback receives an array of relations instead of just a single one.
      See the description of the <xref linkend="sql-truncate"/> statement for
      details.
     </para>
    </sect3>

   </sect2>

   <sect2 id="logicaldecoding-output-plugin-output">
    <title>Functions for Producing Output</title>

    <para>
     To actually produce output, output plugins can write data to
     the <literal>StringInfo</literal> output buffer
     in <literal>ctx-&gt;out</literal> when inside
     the <function>begin_cb</function>, <function>commit_cb</function>,
     or <function>change_cb</function> callbacks. Before writing to the output
     buffer, <function>OutputPluginPrepareWrite(ctx, last_write)</function> has
     to be called, and after finishing writing to the
     buffer, <function>OutputPluginWrite(ctx, last_write)</function> has to be
     called to perform the write. The <parameter>last_write</parameter>
     indicates whether a particular write was the callback's last write.
    </para>

    <para>
     The following example shows how to output data to the consumer of an
     output plugin:
<programlisting>
OutputPluginPrepareWrite(ctx, true);
appendStringInfo(ctx->out, "BEGIN %u", txn->xid);
OutputPluginWrite(ctx, true);
</programlisting>
    </para>
   </sect2>
  </sect1>

  <sect1 id="logicaldecoding-writer">
   <title>Logical Decoding Output Writers</title>

   <para>
    It is possible to add more output methods for logical decoding.
    For details, see
    <filename>src/backend/replication/logical/logicalfuncs.c</filename>.
    Essentially, three functions need to be provided: one to read WAL, one to
    prepare writing output, and one to write the output
    (see <xref linkend="logicaldecoding-output-plugin-output"/>).
   </para>
  </sect1>

  <sect1 id="logicaldecoding-synchronous">
   <title>Synchronous Replication Support for Logical Decoding</title>
   <sect2 id="logicaldecoding-synchronous-overview">
    <title>Overview</title>

    <para>
     Logical decoding can be used to build
     <link linkend="synchronous-replication">synchronous
     replication</link> solutions with the same user interface as synchronous
     replication for <link linkend="streaming-replication">streaming
     replication</link>.  To do this, the streaming replication interface
     (see <xref linkend="logicaldecoding-walsender"/>) must be used to stream out
     data. Clients have to send <literal>Standby status update (F)</literal>
     (see <xref linkend="protocol-replication"/>) messages, just like streaming
     replication clients do.
    </para>

    <note>
     <para>
      A synchronous replica receiving changes via logical decoding will work in
      the scope of a single database. Since, in contrast to
      that, <parameter>synchronous_standby_names</parameter> currently is
      server wide, this means this technique will not work properly if more
      than one database is actively used.
     </para>
    </note>
   </sect2>

   <sect2 id="logicaldecoding-synchronous-caveats">
    <title>Caveats</title>

    <para>
     In synchronous replication setup, a deadlock can happen, if the transaction
     has locked [user] catalog tables exclusively. See
     <xref linkend="logicaldecoding-capabilities"/> for information on user
     catalog tables. This is because logical decoding of transactions can lock
     catalog tables to access them. To avoid this users must refrain from taking
     an exclusive lock on [user] catalog tables. This can happen in the following
     ways:

     <itemizedlist>
      <listitem>
       <para>
        Issuing an explicit <command>LOCK</command> on <structname>pg_class</structname>
        in a transaction.
       </para>
      </listitem>

      <listitem>
       <para>
        Perform <command>CLUSTER</command> on <structname>pg_class</structname> in
        a transaction.
       </para>
      </listitem>

      <listitem>
       <para>
        <command>PREPARE TRANSACTION</command> after <command>LOCK</command> command
        on <structname>pg_class</structname> and allow logical decoding of two-phase
        transactions.
       </para>
      </listitem>

      <listitem>
       <para>
        <command>PREPARE TRANSACTION</command> after <command>CLUSTER</command>
        command on <structname>pg_trigger</structname> and allow logical decoding of
        two-phase transactions. This will lead to deadlock only when published table
        have a trigger.
       </para>
      </listitem>

      <listitem>
       <para>
        Executing <command>TRUNCATE</command> on [user] catalog table in a
        transaction.
       </para>
      </listitem>
     </itemizedlist>

     Note that these commands that can cause deadlock apply to not only explicitly
     indicated system catalog tables above but also to any other [user] catalog
     table.
    </para>
   </sect2>
  </sect1>

  <sect1 id="logicaldecoding-streaming">
   <title>Streaming of Large Transactions for Logical Decoding</title>

   <para>
    The basic output plugin callbacks (e.g., <function>begin_cb</function>,
    <function>change_cb</function>, <function>commit_cb</function> and
    <function>message_cb</function>) are only invoked when the transaction
    actually commits. The changes are still decoded from the transaction
    log, but are only passed to the output plugin at commit (and discarded
    if the transaction aborts).
   </para>

   <para>
    This means that while the decoding happens incrementally, and may spill
    to disk to keep memory usage under control, all the decoded changes have
    to be transmitted when the transaction finally commits (or more precisely,
    when the commit is decoded from the transaction log). Depending on the
    size of the transaction and network bandwidth, the transfer time may
    significantly increase the apply lag.
   </para>

   <para>
    To reduce the apply lag caused by large transactions, an output plugin
    may provide additional callback to support incremental streaming of
    in-progress transactions. There are multiple required streaming callbacks
    (<function>stream_start_cb</function>, <function>stream_stop_cb</function>,
    <function>stream_abort_cb</function>, <function>stream_commit_cb</function>
    and <function>stream_change_cb</function>) and two optional callbacks
    (<function>stream_message_cb</function> and <function>stream_truncate_cb</function>).
    Also, if streaming of two-phase commands is to be supported, then additional
    callbacks must be provided. (See <xref linkend="logicaldecoding-two-phase-commits"/>
    for details).
   </para>

   <para>
    When streaming an in-progress transaction, the changes (and messages) are
    streamed in blocks demarcated by <function>stream_start_cb</function>
    and <function>stream_stop_cb</function> callbacks. Once all the decoded
    changes are transmitted, the transaction can be committed using the
    <function>stream_commit_cb</function> callback
    (or possibly aborted using the <function>stream_abort_cb</function> callback).
    If two-phase commits are supported, the transaction can be prepared using the
    <function>stream_prepare_cb</function> callback,
    <command>COMMIT PREPARED</command> using the
    <function>commit_prepared_cb</function> callback or aborted using the
    <function>rollback_prepared_cb</function>.
   </para>

   <para>
    One example sequence of streaming callback calls for one transaction may
    look like this:
<programlisting>
stream_start_cb(...);   &lt;-- start of first block of changes
  stream_change_cb(...);
  stream_change_cb(...);
  stream_message_cb(...);
  stream_change_cb(...);
  ...
  stream_change_cb(...);
stream_stop_cb(...);    &lt;-- end of first block of changes

stream_start_cb(...);   &lt;-- start of second block of changes
  stream_change_cb(...);
  stream_change_cb(...);
  stream_change_cb(...);
  ...
  stream_message_cb(...);
  stream_change_cb(...);
stream_stop_cb(...);    &lt;-- end of second block of changes


[a. when using normal commit]
stream_commit_cb(...);    &lt;-- commit of the streamed transaction

[b. when using two-phase commit]
stream_prepare_cb(...);   &lt;-- prepare the streamed transaction
commit_prepared_cb(...);  &lt;-- commit of the prepared transaction
</programlisting>
   </para>

   <para>
    The actual sequence of callback calls may be more complicated, of course.
    There may be blocks for multiple streamed transactions, some of the
    transactions may get aborted, etc.
   </para>

   <para>
    Similar to spill-to-disk behavior, streaming is triggered when the total
    amount of changes decoded from the WAL (for all in-progress transactions)
    exceeds the limit defined by <varname>logical_decoding_work_mem</varname> setting.
    At that point, the largest top-level transaction (measured by the amount of memory
    currently used for decoded changes) is selected and streamed.  However, in
    some cases we still have to spill to disk even if streaming is enabled
    because we exceed the memory threshold but still have not decoded the
    complete tuple e.g., only decoded toast table insert but not the main table
    insert.
   </para>

   <para>
    Even when streaming large transactions, the changes are still applied in
    commit order, preserving the same guarantees as the non-streaming mode.
   </para>

  </sect1>

  <sect1 id="logicaldecoding-two-phase-commits">
   <title>Two-phase Commit Support for Logical Decoding</title>

   <para>
    With the basic output plugin callbacks (eg., <function>begin_cb</function>,
    <function>change_cb</function>, <function>commit_cb</function> and
    <function>message_cb</function>) two-phase commit commands like
    <command>PREPARE TRANSACTION</command>, <command>COMMIT PREPARED</command>
    and <command>ROLLBACK PREPARED</command> are not decoded. While the
    <command>PREPARE TRANSACTION</command> is ignored,
    <command>COMMIT PREPARED</command> is decoded as a <command>COMMIT</command>
    and <command>ROLLBACK PREPARED</command> is decoded as a
    <command>ROLLBACK</command>.
   </para>

   <para>
    To support the streaming of two-phase commands, an output plugin needs to
    provide additional callbacks. There are multiple two-phase commit callbacks
    that are required, (<function>begin_prepare_cb</function>,
    <function>prepare_cb</function>, <function>commit_prepared_cb</function>,
    <function>rollback_prepared_cb</function> and
    <function>stream_prepare_cb</function>) and an optional callback
    (<function>filter_prepare_cb</function>).
   </para>

   <para>
    If the output plugin callbacks for decoding two-phase commit commands are
    provided, then on <command>PREPARE TRANSACTION</command>, the changes of
    that transaction are decoded, passed to the output plugin, and the
    <function>prepare_cb</function> callback is invoked. This differs from the
    basic decoding setup where changes are only passed to the output plugin
    when a transaction is committed. The start of a prepared transaction is
    indicated by the <function>begin_prepare_cb</function> callback.
   </para>

   <para>
    When a prepared transaction is rolled back using the
    <command>ROLLBACK PREPARED</command>, then the
    <function>rollback_prepared_cb</function> callback is invoked and when the
    prepared transaction is committed using <command>COMMIT PREPARED</command>,
    then the <function>commit_prepared_cb</function> callback is invoked.
   </para>

   <para>
    Optionally the output plugin can define filtering rules via
    <function>filter_prepare_cb</function> to decode only specific transaction
    in two phases.  This can be achieved by pattern matching on the
    <parameter>gid</parameter> or via lookups using the
    <parameter>xid</parameter>.
   </para>

   <para>
    The users that want to decode prepared transactions need to be careful about
    below mentioned points:

    <itemizedlist>
     <listitem>
      <para>
       If the prepared transaction has locked [user] catalog tables exclusively
       then decoding prepare can block till the main transaction is committed.
      </para>
     </listitem>

     <listitem>
      <para>
       The logical replication solution that builds distributed two phase commit
       using this feature can deadlock if the prepared transaction has locked
       [user] catalog tables exclusively. To avoid this users must refrain from
       having locks on catalog tables (e.g. explicit <command>LOCK</command> command)
       in such transactions.
       See <xref linkend="logicaldecoding-synchronous-caveats"/> for the details.
      </para>
     </listitem>
    </itemizedlist>
   </para>

  </sect1>
 </chapter>