PostgreSQL 9.0 allow us to have replicated Hot Standby servers which we can query and/or use for high availability.

While the main components of the feature are included with PostgreSQL, the user is expected to manage the high availability part of it.

repmgr allows you to monitor and manage your replicated PostgreSQL databases as a single cluster.

repmgr includes two components:

• repmgr: command program that performs tasks and then exits
• repmgrd: management and monitoring daemon that watches the cluster

repmgr is currently aimed for installation on UNIX-like systems that include development tools such as gcc and gmake. It also requires that the rsync utility is available in the PATH of the user running the repmgr programs.

To install and use repmgr and repmgrd follow these steps:

1. Build repmgr programs
2. Set up trusted copy between postgres accounts, needed for the STANDBY CLONE step
3. Check your primary server is correctly configured
4. Write a suitable repmgr.conf for the node
5. Setup repmgrd to aid in failover transitions

Both methods of installation will place the binaries at the same location as your postgres binaries, such as psql. There are two ways to build it. The second requires a full PostgreSQL source code tree to install the program directly into. The first instead uses the PostgreSQL Extension System (PGXS) to install. For this method to work, you will need the pg_config program available in your PATH. In some distributions of PostgreSQL, this requires installing a separate development package in addition to the basic server software.

If you are using a packaged PostgreSQL build and have pg_config available, the package can be built and installed using PGXS instead:

tar xvzf repmgr-1.0.tar.gz
cd repmgr
make USE_PGXS=1
make USE_PGXS=1 install

This is preferred to building from the contrib subdirectory of the main source code tree.

If you need to remove the source code temporary files from this directory, that can be done like this:

make USE_PGXS=1 clean

See below for building notes specific to RedHat Linux variants.

In this method, the repmgr distribution is copied into the PostgreSQL source code tree, assumed to be at the ${postgresql_sources} for this example. The resulting subdirectory must be named contrib/repmgr, without any version number:

cp repmgr.tar.gz ${postgresql_sources}/contrib
cd ${postgresql_sources}/contrib
tar xvzf repmgr-1.0.tar.gz
cd repmgr
make
make install

If you need to remove the source code temporary files from this directory, that can be done like this:

make clean

The RPM packages of PostgreSQL put pg_config into the postgresql-devel package, not the main server one. And if you have a RPM install of PostgreSQL 9.0, the entire PostgreSQL binary directory will not be in your PATH by default either. Individual utilities are made available via the alternatives mechanism, but not all commands will be wrapped that way. The files installed by repmgr will certainly not be in the default PATH for the postgres user on such a system. They will instead be in /usr/pgsql-9.0/bin/ on this type of system.

When building repmgr against a RPM packaged build, you may discover that some development packages are needed as well. The following build errors can occur:

/usr/bin/ld: cannot find -lxslt
/usr/bin/ld: cannot find -lpam

Install the following packages to correct those:

yum install libxslt-devel
yum install pam-devel

If building repmgr as a regular user, then doing the install into the system directories using sudo, the syntax is hard. pg_config won't be in root's path either. The following recipe should work:

sudo PATH="/usr/pgsql-9.0/bin:$PATH" make USE_PGXS=1 install

You should now find the repmgr programs available in the subdirectory where the rest of your PostgreSQL installation is at. You can confirm the software is available by checking its version:

repmgr --version
repmgrd --version

You may need to include the full path of the binary instead, such as this RHEL example:

/usr/pgsql-9.0/bin/repmgr --version
/usr/pgsql-9.0/bin/repmgrd --version

Below this binary installation base directory is referred to as PGDIR.

Initial copy between nodes uses the rsync program running over ssh. For this to work, the postgres accounts on each system need to be able to access files on their partner node without a password.

First generate a ssh key, using an empty passphrase, and copy the resulting keys and a maching authorization file to a privledged user on the other system:

[postgres@db1]$ ssh-keygen -t rsa
Generating public/private rsa key pair.
Enter file in which to save the key (/var/lib/pgsql/.ssh/id_rsa):
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in /var/lib/pgsql/.ssh/id_rsa.
Your public key has been saved in /var/lib/pgsql/.ssh/id_rsa.pub.
The key fingerprint is:
aa:bb:cc:dd:ee:ff:aa:11:22:33:44:55:66:77:88:99 [email protected]
[postgres@db1]$ cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
[postgres@db1]$ chmod go-rwx ~/.ssh/*
[postgres@db1]$ cd ~/.ssh
[postgres@db1]$ scp id_rsa.pub id_rsa authorized_keys user@db2:

Login as a user on the other system, and install the files into the postgres user's account:

[user@db2 ~]$ sudo chown postgres.postgres authorized_keys id_rsa.pub id_rsa
[user@db2 ~]$ sudo mkdir -p ~postgres/.ssh
[user@db2 ~]$ sudo chown postgres.postgres ~postgres/.ssh
[user@db2 ~]$ sudo mv authorized_keys id_rsa.pub id_rsa ~postgres/.ssh
[user@db2 ~]$ sudo chmod -R go-rwx ~postgres/.ssh

Now test that ssh in both directions works. You may have to accept some new known hosts in the process.

PostgreSQL should have been previously built and installed on the system. Here is a sample of changes to the postgresql.conf file:

listen_addresses='*'
wal_level = 'hot_standby'
archive_mode = on
archive_command = 'cd .'       # we can also use exit 0, anything that
                           # just does nothing
max_wal_senders = 10
wal_keep_segments = 5000     # 80 GB required on pg_xlog
hot_standby = on

Also you need to add the machines that will participate in the cluster in pg_hba.conf file. One possibility is to trust all connections from the replication users from all internal addresses, such as:

host     all              all         192.168.1.0/24         trust
host     replication      all         192.168.1.0/24         trust

A more secure setup adds a repmgr user and database, just giving access to that user:

host     repmgr           repmgr      192.168.1.0/24         trust
host     replication      all         192.168.1.0/24         trust

If you give a password to the user, you need to create a .pgpass file for them as well to allow automatic login. In this case you might use the md5 authentication method instead of trust for the repmgr user.

Don't forget to restart the database server after making all these changes.

repmgr.conf is looked for in the directory repmgrd or repmgr exists. The configuration file should have 3 lines:

It should have these three parameters:

1. cluster: A string (single quoted) that identify the cluster we are on
2. node: An integer that identify our node in the cluster
3. conninfo: A string (single quoted) specifying how we can connect to this node's PostgreSQL service

The current supported syntax for the program can be seen using:

repmgr --help

The output from this program looks like this:

repmgr: Replicator manager
Usage:
 repmgr [OPTIONS] master  {register}
 repmgr [OPTIONS] standby {register|clone|promote|follow}

General options:
  --help                     show this help, then exit
  --version                  output version information, then exit
  --verbose                  output verbose activity information

Connection options:
  -d, --dbname=DBNAME        database to connect to
  -h, --host=HOSTNAME        database server host or socket directory
  -p, --port=PORT            database server port
  -U, --username=USERNAME    database user name to connect as

Configuration options:
  -D, --data-dir=DIR         local directory where the files will be copied to
  -f, --config_file=PATH     path to the configuration file
  -R, --remote-user=USERNAME database server username for rsync
  -w, --wal-keep-segments=VALUE  minimum value for the GUC wal_keep_segments (default: 5000)
  -F, --force                force potentially dangerous operations to happen

repmgr performs some tasks like clone a node, promote it or making follow another node and then exits.
COMMANDS:
 master register       - registers the master in a cluster
 standby register      - registers a standby in a cluster
 standby clone [node]  - allows creation of a new standby
 standby promote       - allows manual promotion of a specific standby into a new master in the event of a failover
 standby follow        - allows the standby to re-point itself to a new master

The --verbose option can be useful in troubleshooting issues with the program.

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, it needs to be executed before any node is registered

• standby register

  • Registers a standby in a cluster, it needs to be executed before any repmgrd is executed

• standby clone [node to be cloned]

  • Does a backup via rsync of the data directory of the primary. And it creates the recovery file we need to start a new hot standby server. It doesn't need the repmgr.conf so it can be executed anywhere on the new node. You can change to the directory you want the new database cluster at and execute:

    ./repmgr standby clone 10.68.1.161
    

    or run from wherever you are with a full path:

    ./repmgr -D /path/to/new/data/directory standby clone 10.68.1.161
    

    That will make a backup of the primary then you only need to start the server using a command like:

    pg_ctl -D /your_data_directory_path start
    

    Note that some installations will also redirect the output log file when executing pg_ctl.

• standby promote

  • Allows manual promotion of a specific standby into a new primary in the event of a failover. This needs to be executed on the same directory where the repmgr.conf is in the standby, or you can use the -f option to indicate where the repmgr.conf is at. It doesn't need any additional arguments:

    ./repmgr standby promote
    

    That will restart your standby postgresql service.

• standby follow

  • Allows the standby to base itself to the new primary passed as a parameter. This needs to be executed on the same directory where the repmgr.conf is in the standby, or you can use the -f option to indicate where the repmgr.conf is at. Example:

    ./repmgr standby follow
    

Suppose we have 3 nodes: node1 (the initial master), node2 and node3

To make node2 and node3 be standbys of node1, execute this on both nodes (node2 and node3):

repmgr -D /var/lib/postgresql/9.0 standby clone node1

If we lose node1 we can run on node2:

repmgr -f /home/postgres/repmgr.conf standby promote

Which makes node2 the new master. We then run on node3:

repmgr standby follow

To make node3 follow node2 (rather than node1)

If now we want to add a new node we can a prepare a new server (node4) and run:

repmgr -D /var/lib/postgresql/9.0 standby clone node2

NOTE: you need to have $PGDIR/bin (where the PostgreSQL binaries are installed) in your path for the above to work. If you don't want that as a permanent setting, you can temporarily set it before running individual commands like this:

PATH=$PGDIR/bin:$PATH repmgr standby promote

The current supported syntax for the program can be seen using:

repmgrd --help

The output from this program looks like this:

repmgrd: Replicator manager daemon
Usage:
 repmgrd [OPTIONS]

Options:
  --help                    show this help, then exit
  --version                 output version information, then exit
  --verbose                 output verbose activity information
  -f, --config_file=PATH    database to connect to

repmgrd monitors a cluster of servers.

The --verbose option can be useful in troubleshooting issues with the program.

To use the repmgrd (repmgr daemon) to monitor standby so we know how is going the replication and how far they are from primary, you need to execute the repmgr.sql script in the postgres database.

You also need to add a row for every node in the repl_node table. This work may be done for you by the daemon itself, as described below.

To look at the current lag between primary and each node listed in repl_node, consult the repl_status view:

psql -d postgres -c "SELECT * FROM repl_status"

This view shows the latest monitor info from every node.

• replication_lag: in bytes. This is how far the latest xlog record we have received is from master.
• apply_lag: in bytes. This is how far the latest xlog record we have applied is from the latest record we have received.
• time_lag: in seconds. How many seconds behind the master is this node.

repmgrd reads the repmgr.conf file in current directory, or as indicated with -f parameter. It checks if the standby is in repl_nodes and adds it if not.

Before you can run the repmgr daemon (repmgrd) you need to register a master and at least a standby in a cluster using the MASTER REGISTER and STANDBY REGISTER commands.

For example, following last example and assuming that repmgr.conf is in postgres home directory you will run this on the master:

repmgr -f /home/postgres/repmgr.conf master register

and the same in the standby.

The repmgr daemon creates 2 connections: one to the master and another to the standby.

This assumes you've already followed the steps in "Installation Outline" to install repmgr and repmgr on the system.

The following scenario involves two PostgreSQL installations on the same server hardware, so that additional systems aren't needed for testing. A normal production installation of repmgr will normally involve two different systems running on the same port, typically the default of 5432, with both using files owned by the postgres user account. In places where 127.0.0.1 is used as a host name below, you would instead use the name of the relevant host for that parameter. You can usually leave out changes to the port number in this case too.

The test setup assumes you might be using the default installation of PostgreSQL on port 5432 for some other purpose, and instead relocates these instances onto different ports running as different users:

• A primary (master) server called “prime," with a user as “prime," who is also the owner of the files. This server is operating on port 5433. This server will be known as “node1" in the cluster “test"
• A standby server called “standby", with a user of “standby", who is the owner of the files. This server is operating on port 5434. This server will be known and “node2" on the cluster “test."
• A database exists on “prime" called “testdb."
• The Postgress installation in each of the above is defined as $PGDATA, which is represented here with /data/prime as the "prime" server and /data/standby as the "standby" server.

You might setup such an installation by adjusting the login script for the "prime" and "standby" users as in these two examples:

# prime
PGDATA=/data/prime
PGENGINE=/usr/pgsql-9.0/bin
PGPORT=5433
export PGDATA PGENGINE PGPORT
PATH="$PATH:$PGENGINE"

# standby
PGDATA=/data/standby
PGENGINE=/usr/pgsql-9.0/bin
PGPORT=5434
export PGDATA PGENGINE PGPORT
PATH="$PATH:$PGENGINE"

And then starting/stopping each installation as needed using the pg_ctl utility.

Note: naming your nodes based on their starting role is not a recommended best practice! As you'll see in this example, once there is a failover, names strongly associated with one particular role (primary or standby) can become confusing, once that node no longer has that role. Future versions of this walkthrough are expected to use more generic terminology for these names.

Setup a streaming replica, strip away any PostgreSQL installation on the existing replica:

• Stop both servers.

• Go to “standby" database directory and remove the PostgreSQL installation:

  cd $PGDATA
  rm -rf *
  

  This will delete the entire database installation in /data/standby.

Create a directory to store each repmgr configuration in for each node. In that, there needs to be a repmgr.conf file for each node in the cluster. For “prime" we'll assume this is stored in /home/prime/repmgr and it should contain:

cluster=test
node=1
conninfo='host=127.0.0.1 dbname=dbtest'

On “standby" create the file /home/standby/repmgr/repmgr.conf with:

cluster=test
node=2
conninfo='host=127.0.0.1 dbname=dbtest'

Next, with “prime" server running, we want to use the clone standby command in repmgr to copy over the entire PostgreSQL database cluster onto the “standby" server. On the “standby" server, type:

repmgr -D $PGDATA -p 5433 -U prime -R prime --verbose standby clone localhost

Next, we need a recovery.conf file on “standby" in the $PGDATA directory that reads as follows:

standby_mode = 'on'
primary_conninfo = 'host=127.0.0.1 port=5433'

Make sure that standby has a qualifying role in the database, “testdb" in this case, and can login. Start psql on the testdb database on “prime" and at the testdb# prompt type:

CREATE ROLE standby SUPERUSER LOGIN

First, register the master by typing on “prime":

repmgr -f /home/prime/repmgr/repmgr.conf --verbose master register

On “standby," edit the postgresql.conf file and change the port to 5434.

Start the “standby" server.

Register the standby by typing on “standby":

repmgr -f /home/standby/repmgr/repmgr.conf --verbose standby register

At this point, you have a functioning primary on “prime" and a functioning standby server running on “standby." It's recommended that you insert some records into the primary server here, then confirm they appear very quickly (within milliseconds) on the standby. Also verify that one can make queries against the standby server and cannot make insertions into the standby database.

To simulate the loss of the primary server, simply stop the “prime" server. At this point, the standby contains the database as it existed at the time of the “failure" of the primary server.

Now you can promote the standby server to be the primary, to allow applications to read and write to the database again, by typing:

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

The server restarts and now has read/write ability.

To make the former primary act as a standby, which is necessary before restoring the original roles, type:

repmgr -U standby -R prime -h 127.0.0.1 -p 5433 -d dbtest --force --verbose standby clone

Stop and restart the “prime" server, which is now acting as a standby server.

Make sure the record(s) inserted the earlier step are still available on the now standby (prime). Confirm the database on “prime" is read-only.

Now restore to the original configuration by stopping the “standby" (now acting as a primary), promoting “prime" again to be the primary server, then bringing up “standby" as a standby with a valid recovery.conf file on “standby".

Stop the “standby" server:

repmgr -f /home/prime/repmgr/repmgr.conf standby promote

Now the original primary, “prime" is acting again as primary.

Start the “standby" server and type this on “prime":

repmgr standby clone --force -h 127.0.0.1 -p 5434 -U prime -R standby --verbose

Stop the “standby" and change the port to be 5434 in the postgresql.conf file.

Verify the roles have reversed by attempting to insert a record on “standby" and on “prime."

The servers are now again acting as primary on “prime" and standby on “standby".

repmgr is licensed under the GPL v3. All of its code and documentation is Copyright 2010, 2ndQuadrant Limited. See the files COPYRIGHT and LICENSE for details.

Contributions to repmgr are welcome, and listed in the file CREDITS. 2ndQuadrant Limited requires that any contributions provide a copyright assignment and a disclaimer of any work-for-hire ownership claims from the employer of the developer. This lets us make sure that all of the repmgr distribution remains free code. Please contact [email protected] for a copy of the relevant Copyright Assignment Form.