github/repmgr
1
0
Fork 0
mirror of https://github.com/EnterpriseDB/repmgr.git synced 2026-03-23 15:16:29 +00:00
Code Issues Packages Projects Releases 1 Wiki Activity
Files
a4e79d33afcaf1158dd85d601912a4fa7bc06abf
repmgr/doc/cloning-standbys.sgml
Ian Barwick e7bb3e9d50 Add section on promoting standby
2017-10-05 10:30:29 +09:00

404 lines
19 KiB
Plaintext
Raw Blame History

<chapter id="cloning-standbys" xreflabel="cloning standbys">
<title>Cloning standbys</title>
<para>
</para>
<sect1 id="cloning-from-barman" xreflabel="Cloning from Barman">
<indexterm>
<primary>cloning</primary>
<secondary>from Barman</secondary>
</indexterm>
<title>Cloning a standby from Barman</title>
<para>
<xref linkend="repmgr-standby-clone"> can use
<ulink url="https://www.2ndquadrant.com/">2ndQuadrant</ulink>'s
<ulink url="https://www.pgbarman.org/">Barman</ulink> application
to clone a standby (and also as a fallback source for WAL files).
</para>
<tip>
<simpara>
Barman (aka PgBarman) should be considered as an integral part of any
PostgreSQL replication cluster. For more details see:
<ulink url="https://www.pgbarman.org/">https://www.pgbarman.org/</ulink>.
</simpara>
</tip>
<para>
Barman support provides the following advantages:
<itemizedlist spacing="compact" mark="bullet">
<listitem>
<para>
the primary node does not need to perform a new backup every time a
new standby is cloned
</para>
</listitem>
<listitem>
<para>
a standby node can be disconnected for longer periods without losing
the ability to catch up, and without causing accumulation of WAL
files on the primary node
</para>
</listitem>
<listitem>
<para>
WAL management on the primary becomes much easier as there's no need
to use replication slots, and <varname>wal_keep_segments</varname>
does not need to be set.
</para>
</listitem>
</itemizedlist>
</para>
<sect2 id="cloning-from-barman-prerequisites" xreflabel="Prerequisites for cloning from Barman">
<title>Prerequisites for cloning from Barman</title>
<para>
In order to enable Barman support for <command>repmgr standby clone</command>, following
prerequisites must be met:
<itemizedlist spacing="compact" mark="bullet">
<listitem>
<para>
the <varname>barman_server</varname> setting in <filename>repmgr.conf</filename> is the same as the
server configured in Barman;
</para>
</listitem>
<listitem>
<para>
the <varname>barman_host</varname> setting in <filename>repmgr.conf</filename> is set to the SSH
hostname of the Barman server;
</para>
</listitem>
<listitem>
<para>
the <varname>restore_command</varname> setting in <filename>repmgr.conf</filename> is configured to
use a copy of the <command>barman-wal-restore</command> script shipped with the
<literal>barman-cli</literal> package (see below);
</para>
</listitem>
<listitem>
<para>
the Barman catalogue includes at least one valid backup for this server.
</para>
</listitem>
</itemizedlist>
</para>
<note>
<simpara>
Barman support is automatically enabled if <varname>barman_server</varname>
is set. Normally it is good practice to use Barman, for instance
when fetching a base backup while cloning a standby; in any case,
Barman mode can be disabled using the <literal>--without-barman</literal>
command line option.
</simpara>
</note>
<tip>
<simpara>
If you have a non-default SSH configuration on the Barman
server, e.g. using a port other than 22, then you can set those
parameters in a dedicated Host section in <filename>~/.ssh/config</filename>
corresponding to the value of<varname>barman_host</varname> in
<filename>repmgr.conf</filename>. See the <literal>Host</literal>
section in <command>man 5 ssh_config</command> for more details.
</simpara>
</tip>
<para>
It's now possible to clone a standby from Barman, e.g.:
<programlisting>
NOTICE: using configuration file "/etc/repmgr.conf"
NOTICE: destination directory "/var/lib/postgresql/data" provided
INFO: connecting to Barman server to verify backup for test_cluster
INFO: checking and correcting permissions on existing directory "/var/lib/postgresql/data"
INFO: creating directory "/var/lib/postgresql/data/repmgr"...
INFO: connecting to Barman server to fetch server parameters
INFO: connecting to upstream node
INFO: connected to source node, checking its state
INFO: successfully connected to source node
DETAIL: current installation size is 29 MB
NOTICE: retrieving backup from Barman...
receiving file list ...
(...)
NOTICE: standby clone (from Barman) complete
NOTICE: you can now start your PostgreSQL server
HINT: for example: pg_ctl -D /var/lib/postgresql/data start</programlisting>
</para>
</sect2>
<sect2 id="cloning-from-barman-restore-command" xreflabel="Using Barman as a WAL file source">
<title>Using Barman as a WAL file source</title>
<para>
As a fallback in case streaming replication is interrupted, PostgreSQL can optionally
retrieve WAL files from an archive, such as that provided by Barman. This is done by
setting <varname>restore_command</varname> in <filename>recovery.conf</filename> to
a valid shell command which can retrieve a specified WAL file from the archive.
</para>
<para>
<command>barman-wal-restore</command> is a Python script provided as part of the <literal>barman-cli</literal>
package (Barman 2.0 and later; for Barman 1.x the script is provided separately as
<command>barman-wal-restore.py</command>) which performs this function for Barman.
</para>
<para>
To use <command>barman-wal-restore</command> with &repmgr;
and assuming Barman is located on the <literal>barmansrv</literal> host
and that <command>barman-wal-restore</command> is located as an executable at
<filename>/usr/bin/barman-wal-restore</filename>,
<filename>repmgr.conf</filename> should include the following lines:
<programlisting>
barman_host=barmansrv
barman_server=somedb
restore_command=/usr/bin/barman-wal-restore barmansrv somedb %f %p</programlisting>
</para>
<note>
<simpara>
<command>barman-wal-restore</command> supports command line switches to
control parallelism (<literal>--parallel=N</literal>) and compression (
<literal>--bzip2</literal>, <literal>--gzip</literal>).
</simpara>
</note>
<note>
<para>
To use a non-default Barman configuration file on the Barman server,
specify this in <filename>repmgr.conf</filename> with <filename>barman_config</filename>:
<programlisting>
barman_config=/path/to/barman.conf</programlisting>
</para>
</note>
</sect2>
</sect1>
<sect1 id="cloning-replication-slots" xreflabel="Cloning and replication slots">
<indexterm>
<primary>cloning</primary>
<secondary>replication slots</secondary>
</indexterm>
<indexterm>
<primary>replication slots</primary>
<secondary>cloning</secondary>
</indexterm>
<title>Cloning and replication slots</title>
<para>
Replication slots were introduced with PostgreSQL 9.4 and are designed to ensure
that any standby connected to the primary using a replication slot will always
be able to retrieve the required WAL files. This removes the need to manually
manage WAL file retention by estimating the number of WAL files that need to
be maintained on the primary using <varname>wal_keep_segments</varname>.
Do however be aware that if a standby is disconnected, WAL will continue to
accumulate on the primary until either the standby reconnects or the replication
slot is dropped.
</para>
<para>
To enable &repmgr; to use replication slots, set the boolean parameter
<varname>use_replication_slots</varname> in <filename>repmgr.conf</filename>:
<programlisting>
use_replication_slots=true
</programlisting>
</para>
<para>
Replication slots must be enabled in <filename>postgresql.conf</filename>` by
setting the parameter <varname>max_replication_slots</varname> to at least the
number of expected standbys (changes to this parameter require a server restart).
</para>
<para>
When cloning a standby, &repmgr; will automatically generate an appropriate
slot name, which is stored in the <literal>repmgr.nodes</literal> table, and create the slot
on the upstream node:
<programlisting>
repmgr=# SELECT node_id, upstream_node_id, active, node_name, type, priority, slot_name
FROM repmgr.nodes ORDER BY node_id;
node_id | upstream_node_id | active | node_name | type | priority | slot_name
---------+------------------+--------+-----------+---------+----------+---------------
1 | | t | node1 | primary | 100 | repmgr_slot_1
2 | 1 | t | node2 | standby | 100 | repmgr_slot_2
3 | 1 | t | node3 | standby | 100 | repmgr_slot_3
(3 rows)</programlisting>
<programlisting>
repmgr=# SELECT slot_name, slot_type, active, active_pid FROM pg_replication_slots ;
slot_name | slot_type | active | active_pid
---------------+-----------+--------+------------
repmgr_slot_2 | physical | t | 23658
repmgr_slot_3 | physical | t | 23687
(2 rows)</programlisting>
</para>
<para>
Note that a slot name will be created by default for the primary but not
actually used unless the primary is converted to a standby using e.g.
<command>repmgr standby switchover</command>.
</para>
<para>
Further information on replication slots in the PostgreSQL documentation:
<ulink url="https://www.postgresql.org/docs/current/interactive/warm-standby.html#STREAMING-REPLICATION-SLOTS">https://www.postgresql.org/docs/current/interactive/warm-standby.html#STREAMING-REPLICATION-SLOTS</ulink>
</para>
<tip>
<simpara>
While replication slots can be useful for streaming replication, it's
recommended to monitor for inactive slots as these will cause WAL files to
build up indefinitely, possibly leading to server failure.
</simpara>
<simpara>
As an alternative we recommend using 2ndQuadrant's <ulink url="https://www.pgbarman.org/">Barman</ulink>,
which offloads WAL management to a separate server, negating the need to use replication
slots to reserve WAL. See section <xref linkend="cloning-from-barman">
for more details on using &repmgr; together with Barman.
</simpara>
</tip>
</sect1>
<sect1 id="cloning-cascading" xreflabel="Cloning and cascading replication">
<indexterm>
<primary>cloning</primary>
<secondary>cascading replication</secondary>
</indexterm>
<title>Cloning and cascading replication</title>
<para>
Cascading replication, introduced with PostgreSQL 9.2, enables a standby server
to replicate from another standby server rather than directly from the primary,
meaning replication changes "cascade" down through a hierarchy of servers. This
can be used to reduce load on the primary and minimize bandwith usage between
sites. For more details, see the
<ulink url="https://www.postgresql.org/docs/current/static/warm-standby.html#CASCADING-REPLICATION">
PostgreSQL cascading replication documentation</ulink>.
</para>
<para>
&repmgr; supports cascading replication. When cloning a standby,
set the command-line parameter <literal>--upstream-node-id</literal> to the
<varname>node_id</varname> of the server the standby should connect to, and
&repmgr; will create <filename>recovery.conf</filename> to point to it. Note
that if <literal>--upstream-node-id</literal> is not explicitly provided,
&repmgr; will set the standby's <filename>recovery.conf</filename> to
point to the primary node.
</para>
<para>
To demonstrate cascading replication, first ensure you have a primary and standby
set up as shown in the <xref linkend="quickstart">.
Then create an additional standby server with <filename>repmgr.conf</filename> looking
like this:
<programlisting>
node_id=3
node_name=node3
conninfo='host=node3 user=repmgr dbname=repmgr'
data_directory='/var/lib/postgresql/data'</programlisting>
</para>
<para>
Clone this standby (using the connection parameters for the existing standby),
ensuring <literal>--upstream-node-id</literal> is provide with the <varname>node_id</varname>
of the previously created standby (if following the example, this will be <literal>2</literal>):
<programlisting>
$ repmgr -h node2 -U repmgr -d repmgr -f /etc/repmgr.conf standby clone --upstream-node-id=2
NOTICE: using configuration file "/etc/repmgr.conf"
NOTICE: destination directory "/var/lib/postgresql/data" provided
INFO: connecting to upstream node
INFO: connected to source node, checking its state
NOTICE: checking for available walsenders on upstream node (2 required)
INFO: sufficient walsenders available on upstream node (2 required)
INFO: successfully connected to source node
DETAIL: current installation size is 29 MB
INFO: creating directory "/var/lib/postgresql/data"...
NOTICE: starting backup (using pg_basebackup)...
HINT: this may take some time; consider using the -c/--fast-checkpoint option
INFO: executing: 'pg_basebackup -l "repmgr base backup" -D /var/lib/postgresql/data -h node2 -U repmgr -X stream '
NOTICE: standby clone (using pg_basebackup) complete
NOTICE: you can now start your PostgreSQL server
HINT: for example: pg_ctl -D /var/lib/postgresql/data start</programlisting>
then register it (note that <literal>--upstream-node-id</literal> must be provided here
too):
<programlisting>
$ repmgr -f /etc/repmgr.conf standby register --upstream-node-id=2
NOTICE: standby node "node2" (ID: 2) successfully registered
</programlisting>
</para>
<para>
After starting the standby, the cluster will look like this, showing that <literal>node3</literal>
is attached to <literal>node2</literal>, not the primary (<literal>node1</literal>).
<programlisting>
$ repmgr -f /etc/repmgr.conf cluster show
ID | Name | Role | Status | Upstream | Location | Connection string
----+-------+---------+-----------+----------+----------+--------------------------------------
1 | node1 | primary | * running | | default | host=node1 dbname=repmgr user=repmgr
2 | node2 | standby | running | node1 | default | host=node2 dbname=repmgr user=repmgr
3 | node3 | standby | running | node2 | default | host=node3 dbname=repmgr user=repmgr
</programlisting>
</para>
<tip>
<simpara>
Under some circumstances when setting up a cascading replication
cluster, you may wish to clone a downstream standby whose upstream node
does not yet exist. In this case you can clone from the primary (or
another upstream node); provide the parameter <literal>--upstream-conninfo</literal>
to explictly set the upstream's <varname>primary_conninfo</varname> string
in <filename>recovery.conf</filename>.
</simpara>
</tip>
</sect1>
<sect1 id="cloning-advanced" xreflabel="Advanced cloning options">
<indexterm>
<primary>cloning</primary>
<secondary>advanced options</secondary>
</indexterm>
<title>Advanced cloning options</title>
<sect2 id="cloning-advanced-pg-basebackup-options" xreflabel="pg_basebackup options when cloning a standby">
<title>pg_basebackup options when cloning a standby</title>
<para>
By default, <command>pg_basebackup</command> performs a checkpoint before beginning the backup
process. However, a normal checkpoint may take some time to complete;
a fast checkpoint can be forced with the <literal>-c/--fast-checkpoint</literal> option.
However this may impact performance of the server being cloned from
so should be used with care.
</para>
<para>
Further options can be passed to the <command>pg_basebackup</command> utility via
the setting <varname>pg_basebackup_options</varname> in <filename>repmgr.conf</filename>.
See the <ulink url="https://www.postgresql.org/docs/current/static/app-pgbasebackup.html">PostgreSQL pg_basebackup documentation</ulink>
for more details of available options.
</para>
</sect2>
<sect2 id="cloning-advanced-managing-passwords" xreflabel="Managing passwords">
<title>Managing passwords</title>
<para>
If replication connections to a standby's upstream server are password-protected,
the standby must be able to provide the password so it can begin streaming
replication.
</para>
<para>
The recommended way to do this is to store the password in the <literal>postgres</literal> system
user's <filename>~/.pgpass</filename> file. It's also possible to store the password in the
environment variable <varname>PGPASSWORD</varname>, however this is not recommended for
security reasons. For more details see the
<ulink url="https://www.postgresql.org/docs/current/static/libpq-pgpass.html">PostgreSQL password file documentation</ulink>.
</para>
<para>
If, for whatever reason, you wish to include the password in <filename>recovery.conf</filename>,
set <varname>use_primary_conninfo_password</varname> to <literal>true</literal> in
<filename>repmgr.conf</filename>. This will read a password set in <varname>PGPASSWORD</varname>
(but not <filename>~/.pgpass</filename>) and place it into the <literal>primary_conninfo</literal>
string in <filename>recovery.conf</filename>. Note that <varname>PGPASSWORD</varname>
will need to be set during any action which causes <filename>recovery.conf</filename> to be
rewritten, e.g. <xref linkend="repmgr-standby-follow">.
</para>
<para>
It is of course also possible to include the password value in the <varname>conninfo</varname>
string for each node, but this is obviously a security risk and should be
avoided.
</para>
</sect2>
<sect2 id="cloning-advanced-replication-user" xreflabel="Separate replication user">
<title>Separate replication user</title>
<para>
In some circumstances it might be desirable to create a dedicated replication-only
user (in addition to the user who manages the &repmgr; metadata). In this case,
the replication user should be set in <filename>repmgr.conf</filename> via the parameter
<varname>replication_user</varname>; &repmgr; will use this value when making
replication connections and generating <filename>recovery.conf</filename>. This
value will also be stored in the <literal>repmgr.nodes</literal>
table for each node; it no longer needs to be explicitly specified when
cloning a node or executing <xref linkend="repmgr-standby-follow">.
</para>
</sect2>
</sect1>
</chapter>
Reference in New Issue View Git Blame Copy Permalink