repmgr is an open-source tool to manage replication and failover between multiple PostgreSQL servers. It enhances PostgreSQL's built-in hot-standby capabilities with tools to set up standby servers, monitor replication, and perform administrative tasks such as failover or manual switchover operations.

This document covers repmgr 3, which supports PostgreSQL 9.4 and 9.3. This version can use pg_basebackup to clone standby servers, supports replication slots and cascading replication, doesn't require a restart after promotion, and has many usability improvements.

Please continue to use repmgr 2 with earlier PostgreSQL 9.x versions. For a list of changes since repmgr 2 and instructions on upgrading to repmgr 3, see the "Upgrading from repmgr 2" section below.

The repmgr command-line tool is used to perform administrative tasks, and the repmgrd daemon is used to optionally monitor replication and manage automatic failover.

To get started, each PostgreSQL node in your cluster must have a repmgr.conf file. The current master node must be registered using repmgr master register. Existing standby servers can be registered using repmgr standby register. A new standby server can be created using repmgr standby clone followed by repmgr standby register.

See the QUICKSTART.md file for examples of how to use these commands.

Once the cluster is in operation, run repmgr cluster show to see the status of the registered primary and standby nodes. Any standby can be manually promoted using repmgr standby promote. Other standby nodes can be told to follow the new master using repmgr standby follow. We show examples of these commands below.

Next, for detailed monitoring, you must run repmgrd (with the same configuration file) on all your nodes. Replication status information is stored in a custom schema along with information about registered nodes. You also need repmgrd to configure automatic failover in your cluster.

See the FAILOVER.rst file for an explanation of how to set up automatic failover.

repmgr is developed and tested on Linux and OS X, but it should work on any UNIX-like system which PostgreSQL itself supports.

All nodes must be running the same major version of PostgreSQL, and we recommend that they also run the same minor version. This version of repmgr (v3) supports PostgreSQL 9.3 and 9.4.

Earlier versions of repmgr needed password-less SSH access between nodes in order to clone standby servers using rsync. repmgr 3 can use pg_basebackup instead in most circumstances; ssh is not required.

You will need to use rsync only if your PostgreSQL configuration files are outside your data directory (as on Debian) and you wish these to be copied by repmgr. See the SSH-RSYNC.md file for details on configuring password-less SSH between your nodes.

repmgr must be installed on each PostgreSQL server node.

• Packages

  • PGDG publishes RPM packages for RedHat-based distributions
  • Debian/Ubuntu provide .deb packages.
  • See PACKAGES.md for details on building .deb and .rpm packages from the repmgr source code.

• Source installation

  • git clone https://github.com/2ndQuadrant/repmgr
  • Or download tar.gz files from https://github.com/2ndQuadrant/repmgr/releases
  • To install from source, run sudo make USE_PGXS=1 install

After installation, you should be able to run repmgr --version and repmgrd --version. These binaries should be installed in the same directory as other PostgreSQL binaries, such as psql.

By default, repmgr uses PostgreSQL's built-in replication protocol to clone a primary and create a standby server. If your configuration files live outside your data directory, however, you will still need to set up password-less SSH so that rsync can be used. See the SSH-RSYNC.md file for details.

The primary server needs to be configured for replication with settings like the followingin postgresql.conf:

# Allow read-only queries on standby servers. The number of WAL
# senders should be larger than the number of standby servers.

hot_standby = on
wal_level = 'hot_standby'
max_wal_senders = 10

# How much WAL to retain on the primary to allow a temporarily
# disconnected standby to catch up again. The larger this is, the
# longer the standby can be disconnected. This is needed only in
# 9.3; in 9.4, replication slots can be used instead (see below).

wal_keep_segments = 5000

# Enable archiving, but leave it unconfigured (so that it can be
# configured without a restart later). Recommended, not required.

archive_mode = on
archive_command = 'cd .'

PostgreSQL 9.4 makes it possible to use replication slots, which means the value of wal_keep_segments need no longer be set. See section "Replication slots" below for more details.

With PostgreSQL 9.3, repmgr expects wal_keep_segments to be set to at least 5000 (= 80GB of WAL) by default, though this can be overriden with the -w N argument.

A dedicated PostgreSQL superuser account and a database in which to store monitoring and replication data are required. Create them by running the following commands:

createuser -s repmgr
createdb repmgr -O repmgr

We recommend using the name repmgr for both, but you can use whatever name you like (and you need to set the names you chose in the conninfo string in repmgr.conf; see below). repmgr will create the schema and objects it needs when it connects to the server.

Create a repmgr.conf file on each server. Here's a minimal sample:

cluster=test
node=1
node_name=node1
conninfo='host=repmgr_node1 user=repmgr dbname=repmgr'

The cluster name must be the same on all nodes. The node (an integer) and node_name must be unique to each node.

The conninfo string must point to repmgr's database on this node. The host must be an IP or a name that all the nodes in the cluster can resolve (not localhost!). All nodes must use the same username and database name, but other parameters, such as the port, can vary between nodes.

Your repmgr.conf should not be stored inside the PostgreSQL data directory. We recommend /etc/repmgr/repmgr.conf, but you can place it anywhere and use the -f /path/to/repmgr.conf option to tell repmgr where it is. If not specified, repmgr will search for repmgr.conf in the current working directory.

If your PostgreSQL binaries (pg_ctl, pg_basebackup) are not in your PATH, you can specify an alternate location in repmgr.conf:

pg_bindir=/path/to/postgres/bin

See repmgr.conf.sample for an example configuration file with all available configuration settings annotated.

The master node must be registered first using repmgr master register, and standby servers must be registered using repmgr standby register; this inserts details about each node into the control database. Use repmgr cluster show to see the result.

See the QUICKSTART.md file for examples of how to use these commands.

To promote a standby to master, on the standby execute e.g.:

repmgr -f /etc/repmgr/repmgr.conf --verbose standby promote

repmgr will attempt to connect to the current master to verify that it is not available (if it is, repmgr will not promote the standby).

Other standby servers need to be told to follow the new master with e.g.:

repmgr -f /etc/repmgr/repmgr.conf --verbose standby follow

See file FAILOVER.rst for details on setting up automated failover.

Often it's desirable to bring a failed master back into replication as a standby. First, ensure that the master's PostgreSQL server is no longer running; then use repmgr standby clone to re-sync its data directory with the current master, e.g.:

repmgr -f /etc/repmgr/repmgr.conf \
  --force --rsync-only \
  -h node2 -d repmgr -U repmgr --verbose \
  standby clone

Here it's essential to use the command line options --force, to ensure repmgr will re-use the existing data directory, and --rsync-only, which causes repmgr to use rsync rather than pg_basebackup, as the latter can only be used to clone a fresh standby.

The node can then be restarted.

The node will then need to be re-registered with repmgr; again the --force option is required to update the existing record:

 repmgr -f /etc/repmgr/repmgr.conf
   --force \
   standby register

repmgrd is a management and monitoring daemon which runs on standby nodes and which can automate actions such as failover and updating standbys to follow the new master.repmgrd can be started simply with e.g.:

repmgrd -f /etc/repmgr/repmgr.conf --verbose > $HOME/repmgr/repmgr.log 2>&1

or alternatively:

repmgrd -f /etc/repmgr/repmgr.conf --verbose --monitoring-history > $HOME/repmgr/repmgrd.log 2>&1

which will track replication advance or lag on all registered standbys.

For permanent operation, we recommend using the options -d/--daemonize to detach the repmgrd process, and -p/--pid-file to write the process PID to a file.

Example log output (at default log level):

[2015-03-11 13:15:40] [INFO] checking cluster configuration with schema 'repmgr_test'
[2015-03-11 13:15:40] [INFO] checking node 2 in cluster 'test'
[2015-03-11 13:15:40] [INFO] reloading configuration file and updating repmgr tables
[2015-03-11 13:15:40] [INFO] starting continuous standby node monitoring

In a situation caused e.g. by a network interruption between two data centres, it's important to avoid a "split-brain" situation where both sides of the network assume they are the active segment and the side without an active master unilaterally promotes one of its standbys.

To prevent this situation happening, it's essential to ensure that one network segment has a "voting majority", so other segments will know they're in the minority and not attempt to promote a new master. Where an odd number of servers exists, this is not an issue. However, if each network has an even number of nodes, it's necessary to provide some way of ensuring a majority, which is where the witness server becomes useful.

This is not a fully-fledged standby node and is not integrated into replication, but it effectively represents the "casting vote" when deciding which network segment has a majority. A witness server can be set up using repmgr witness create (see below for details) and can run on a dedicated server or an existing node. Note that it only makes sense to create a witness server in conjunction with running repmgrd; the witness server will require its own repmgrd instance.

When repmgrd is running with the option -m/--monitoring-history, it will constantly write node status information to the repl_monitor table, which can be queried easily using the view repl_status:

repmgr=# SELECT * FROM repmgr_test.repl_status;
-[ RECORD 1 ]-------------+-----------------------------
primary_node              | 1
standby_node              | 2
standby_name              | node2
node_type                 | standby
active                    | t
last_monitor_time         | 2015-03-11 14:02:34.51713+09
last_wal_primary_location | 0/3012AF0
last_wal_standby_location | 0/3012AF0
replication_lag           | 0 bytes
replication_time_lag      | 00:00:03.463085
apply_lag                 | 0 bytes
communication_time_lag    | 00:00:00.955385

To help understand what significant events (e.g. failure of a node) happened when and for what reason, repmgr logs such events into the repl_events table, e.g.:

repmgr_db=# SELECT * from repmgr_test.repl_events ;
 node_id |      event       | successful |        event_timestamp        |                                      details
---------+------------------+------------+-------------------------------+-----------------------------------------------------------------------------------
       1 | master_register  | t          | 2015-03-16 17:36:21.711796+09 |
       2 | standby_clone    | t          | 2015-03-16 17:36:31.286934+09 | Cloned from host 'localhost', port 5500; backup method: pg_basebackup; --force: N
       2 | standby_register | t          | 2015-03-16 17:36:32.391567+09 |
(3 rows)

Additionally repmgr can execute an external program each time an event is logged. This program is defined with the configuration variable event_notification_command; the command string can contain the following placeholders, which will be replaced with the same content which is written to the repl_events table:

%n - node id
%e - event type
%s - success (1 or 0)
%t - timestamp
%d - description

Example:

event_notification_command=/path/to/some-script %n %e %s "%t" "%d"

By default the program defined with event_notification_command will be executed for every event; to restrict execution to certain events, list these in the parameter event_notifications

event_notifications=master_register,standby_register

Following event types currently exist:

master_register
standby_register
standby_clone
standby_promote
witness_create
repmgrd_start
repmgrd_failover_promote
repmgrd_failover_follow

Cascading replication - where a standby can connect to an upstream node and not the master server itself - was introduced in PostgreSQL 9.2. repmgr and repmgrd support cascading replication by keeping track of the relationship between standby servers - each node record is stored with the node id of its upstream ("parent") server (except of course the master server).

In a failover situation where the master node fails and a top-level standby is promoted, a standby connected to another standby will not be affected and continue working as normal (even if the upstream standby it's connected to becomes the master node). If however the node's direct upstream fails, the "cascaded standby" will attempt to reconnect to that node's parent.

To configure standby servers for cascading replication, add the parameter upstream_node to repmgr.conf and set it to the id of the node it should connect to, e.g.:

cluster=test
node=2
node_name=node2
upstream_node=1

Replication slots were introduced with PostgreSQL 9.4 and enable standbys to notify the master of their WAL consumption, ensuring that the master will not remove any WAL files until they have been received by all standbys. This mitigates the requirement to manage WAL file retention using wal_keep_segments etc., with the caveat that if a standby fails, no WAL files will be removed until the standby's replication slot is deleted.

To enable replication slots, set the boolean parameter use_replication_slots in repmgr.conf:

use_replication_slots=1

repmgr will automatically generate an appropriate slot name, which is stored in the repl_nodes table.

Note that repmgr will fail with an error if this option is specified when working with PostgreSQL 9.3.

Be aware that when initially cloning a standby, you will need to ensure that all required WAL files remain available while the cloning is taking place. If using the default pg_basebackup method, we recommend setting pg_basebackup's --xlog-method parameter to stream like this:

pg_basebackup_options='--xlog-method=stream'

See the pg_basebackup documentation [*] for details. Otherwise you'll need to set wal_keep_segments to an appropriately high value.

[*] http://www.postgresql.org/docs/current/static/app-pgbasebackup.html

Further reading:

• http://www.postgresql.org/docs/current/interactive/warm-standby.html#STREAMING-REPLICATION-SLOTS
• http://blog.2ndquadrant.com/postgresql-9-4-slots/

repmgr 3 is largely compatible with repmgr 2; the only step required to upgrade is to update the repl_nodes table to the definition needed by repmgr 3. See the file sql/repmgr2_repmgr3.sql for details on how to do this.

repmgrd must not be running while repl_nodes is being updated.

Existing repmgr.conf files can be retained as-is.

Not all of these commands need the repmgr.conf file, but they need to be able to connect to the remote and local databases.

You can teach it which is the remote database by using the -h parameter or as a last parameter in standby clone and standby follow. If you need to specify a port different then the default 5432 you can specify a -p parameter. Standby is always considered as localhost and a second -p parameter will indicate its port if is different from the default one.

• master register

  Registers a master in a cluster. This command needs to be executed before any standby nodes are registered.

• standby register

  Registers a standby with repmgr. This command needs to be executed to enable promote/follow operations and to allow repmgrd to work with the node. An existing standby can be registered using this command.

• standby clone [node to be cloned]

  Clones a new standby node from the data directory of the master (or an upstream cascading standby) using pg_basebackup or rsync. Additionally it will create the recovery.conf file required to start the server as a standby. This command does not require repmgr.conf to be provided, but does require connection details of the master or upstream server as command line parameters.

  Provide the -D/--data-dir option to specify the destination data directory; if not, the same directory path as on the source server will be used. By default, pg_basebackup will be used to copy data from the master or upstream node but this can only be used for bootstrapping new installations. To update an existing but 'stale' data directory (for example belonging to a failed master), rsync must be used by specifying --rsync-only. In this case, password-less SSH connections between servers are required.

• standby promote

  Promotes a standby to a master if the current master has failed. This command requires a valid repmgr.conf file for the standby, either specified explicitly with -f/--config-file or located in the current working directory; no additional arguments are required.

  If the standby promotion succeeds, the server will not need to be restarted. However any other standbys will need to follow the new server, by using standby follow (see below); if repmgrd is active, it will handle this.

  This command will not function if the current master is still running.

• witness create

  Creates a witness server as a separate PostgreSQL instance. This instance can be on a separate server or a server running an existing node. The witness server contain a copy of the repmgr metadata tables but will not be set up as a standby; instead it will update its metadata copy each time a failover occurs.

  Note that it only makes sense to create a witness server if repmgrd is in use; see section "witness server" above.

  By default the witness server will use port 5499 to facilitate easier setup on a server running an existing node.

• standby follow

  Attaches the standby to a new master. This command requires a valid repmgr.conf file for the standby, either specified explicitly with -f/--config-file or located in the current working directory; no additional arguments are required.

  This command will force a restart of the standby server. It can only be used to attach a standby to a new master node.

• cluster show

  Displays information about each node in the replication cluster. This command polls each registered server and shows its role (master / standby / witness) or "FAILED" if the node doesn't respond. It polls each server directly and can be run on any node in the cluster; this is also useful when analyzing connectivity from a particular node.

  This command requires a valid repmgr.conf file for the node on which it is executed, either specified explicitly with -f/--config-file or located in the current working directory; no additional arguments are required.

  Example:

    repmgr -f /path/to/repmgr.conf cluster show
    Role      | Connection String
    * master  | host=node1 dbname=repmgr user=repmgr
      standby | host=node2 dbname=repmgr user=repmgr
      standby | host=node3 dbname=repmgr user=repmgr
  

• cluster cleanup

  Purges monitoring history from the repl_monitor table to prevent excessive table growth. Use the -k/--keep-history to specify the number of days of monitoring history to retain. This command can be used manually or as a cronjob.

  This command requires a valid repmgr.conf file for the node on which it is executed, either specified explicitly with -f/--config-file or located in the current working directory; no additional arguments are required.

See repmgr.conf.sample for an example configuration file with available configuration settings annotated.

repmgr creates a small schema for its own use in the database specified in each node's conninfo configuration parameter. This database can in principle be any database. The schema name is the global cluster name prefixed with repmgr_, so for the example setup above the schema name is repmgr_test.

The schema contains two tables:

• repl_nodes stores information about all registered servers in the cluster
• repl_monitor stores monitoring information about each node (generated by repmgrd with -m/--monitoring-history option enabled)

and one view:

• repl_status summarizes the latest monitoring information for each node (generated by repmgrd with -m/--monitoring-history option enabled)

repmgr or repmgrd will return one of the following error codes on program exit:

• SUCCESS (0) Program ran successfully.
• ERR_BAD_CONFIG (1) Configuration file could not be parsed or was invalid
• ERR_BAD_RSYNC (2) An rsync call made by the program returned an error
• ERR_NO_RESTART (4) An attempt to restart a PostgreSQL instance failed
• ERR_DB_CON (6) Error when trying to connect to a database
• ERR_DB_QUERY (7) Error while executing a database query
• ERR_PROMOTED (8) Exiting program because the node has been promoted to master
• ERR_BAD_PASSWORD (9) Password used to connect to a database was rejected
• ERR_STR_OVERFLOW (10) String overflow error
• ERR_FAILOVER_FAIL (11) Error encountered during failover (repmgrd only)
• ERR_BAD_SSH (12) Error when connecting to remote host via SSH
• ERR_SYS_FAILURE (13) Error when forking (repmgrd only)
• ERR_BAD_BASEBACKUP (14) Error when executing pg_basebackup

2ndQuadrant provides 24x7 production support for repmgr, including configuration assistance, installation verification and training for running a robust replication cluster. For further details see:

• http://2ndquadrant.com/en/support/

There is a mailing list/forum to discuss contributions or issues http://groups.google.com/group/repmgr

The IRC channel #repmgr is registered with freenode.

Further information is available at http://www.repmgr.org/

We'd love to hear from you about how you use repmgr. Case studies and news are always welcome. Send us an email at [email protected], or send a postcard to

repmgr
c/o 2ndQuadrant
7200 The Quorum
Oxford Business Park North
Oxford
OX4 2JZ
United Kingdom

Thanks from the repmgr core team.

• Ian Barwick
• Jaime Casanova
• Abhijit Menon-Sen
• Simon Riggs
• Cedric Villemain

• http://blog.2ndquadrant.com/announcing-repmgr-2-0/
• http://blog.2ndquadrant.com/managing-useful-clusters-repmgr/
• http://blog.2ndquadrant.com/easier_postgresql_90_clusters/