Navigation :

pgpool-II 4.0.4 Documentation > Pgpool-II + Watchdog Setup Example

Pgpool-II + Watchdog Setup Example

pgpool-II 4.0.4 Documentation
Prev	Up	Chapter 7. Configuration Examples	Next

This section shows an example of streaming replication configuration using Pgpool-II . In this example, we use 3 Pgpool-II servers to manage PostgreSQL servers to create a robust cluster system and avoid the single point of failure or split brain.

7.3.1. Requirements

We assume that all the Pgpool-II servers and the PostgreSQL servers are in the same subnet.

7.3.2. Cluster System Configuration

We use 3 servers with CentOS 7.4. Let these servers be server1 , server2 , server3 . We install PostgreSQL and Pgpool-II on each server.

Figure 7-1. Cluster System Configuration

Note: The roles of Active , Standy , Primary , Standby are not fixed and may be changed by further operations.

Table 7-2. Hostname and IP address

Hostname	IP Address	Virtual IP
server1	192.168.137.101	192.168.137.150
server2	192.168.137.102
server3	192.168.137.103

Table 7-3. PostgreSQL version and Configuration

Item	Value	Detail
PostgreSQL Version	11.1	-
port	5432	-
$PGDATA	/var/lib/pgsql/11/data	-
Archive mode	on	/var/lib/pgsql/archivedir
Start automatically	Disable	-

Table 7-4. Pgpool-II version and Configuration

Item	Value	Detail
Pgpool-II Version	4.0.2	-
port	9999	Pgpool-II accepts connections
	9898	PCP process accepts connections
	9000	watchdog accepts connections
	9694	UDP port for receiving Watchdog's heartbeat signal
Config file	/etc/pgpool-II/pgpool.conf	Pgpool-II config file
Pgpool-II start user	root	See Section 2.1.7 to startup Pgpool-II with non-root user
Running mode	streaming replication mode	-
Watchdog	on	Life check method: heartbeat
Start automatically	Disable	-

7.3.3. Installation

In this example, we install Pgpool-II 4.0.2 and PostgreSQL 11.1 by using RPM packages.

Install PostgreSQL by using PostgreSQL YUM repository.

# yum install https://download.postgresql.org/pub/repos/yum/11/redhat/rhel-7-x86_64/pgdg-centos11-11-2.noarch.rpm
# yum install postgresql11 postgresql11-libs postgresql11-devel postgresql11-server

Install Pgpool-II by using Pgpool-II YUM repository.

# yum install http://www.pgpool.net/yum/rpms/4.0/redhat/rhel-7-x86_64/pgpool-II-release-4.0-1.noarch.rpm
# yum install pgpool-II-pg11-*

7.3.4. Before Starting

Before you start the configuration process, please check the following prerequisites.

Set up PostgreSQL streaming replication on the primary server. In this example, we use WAL archiving.

First, we create the directory /var/lib/pgsql/archivedir to store WAL segments on all servers.
```
[all servers]# su - postgres
[all servers]$ mkdir /var/lib/pgsql/archivedir
          
```
Then we edit the configuration file $PGDATA/postgresql.conf on server1 (primary) as follows.
```
listen_addresses = '*'
archive_mode = on
archive_command = 'cp "%p" "/var/lib/pgsql/archivedir/%f"'
          
```
We use the online recovery functionality of Pgpool-II to setup standby server after the primary server is started.

Because of the security reasons, we create a user repl solely used for replication purpose, and a user pgpool for streaming replication delay check and health check of Pgpool-II .

Table 7-5. Users

User Name	Passowrd	Detail
repl	repl	PostgreSQL replication user
pgpool	pgpool	Pgpool-II health check and replication delay check user
postgres	postgres	User running online recovery

[server1]# psql -U postgres -p 5432
postgres=# SET password_encryption = 'scram-sha-256';
postgres=# CREATE ROLE pgpool WITH LOGIN;
postgres=# CREATE ROLE repl WITH REPLICATION;
postgres=# \password pgpool
postgres=# \password repl
postgres=# \password postgres

Note: If you plan to use detach_false_primary , role "pgpool" needs to be PostgreSQL super user or or in "pg_monitor" group to use this feature. To let "pgpool" role be in the group, you can do:
GRANT pg_monitor TO pgpool;
	  

Assuming that all the Pgpool-II servers and the PostgreSQL servers are in the network of 192.168.137.0/24 , and edit pg_hba.conf to enable scram-sha-256 authentication method.

host    all             pgpool          192.168.137.0/24         scram-sha-256
host    all             all             0.0.0.0/0                scram-sha-256

host    replication     repl            192.168.137.0/24         scram-sha-256

To use the failover and online recovery of Pgpool-II , the settings that allow SSH without passowrd to other servers are necessary.

To allow repl user without specifying password for streaming replication and online recovery, we create the .pgpass file in postgres user's home directory and change the permisson to 600 on each PostgreSQL server.

[all servers]# su - postgres
[all servers]$ vi /var/lib/pgsql/.pgpass
server1:5432:replication:repl:<repl user password>
server2:5432:replication:repl:<repl user passowrd>
server3:5432:replication:repl:<repl user passowrd>
[all servers]$ chmod 600  /var/lib/pgsql/.pgpass

When connect to Pgpool-II and PostgreSQL servers, the target port must be accessible by enabling firewall management softwares. Following is an example for CentOS/RHEL7 .

[all servers]# firewall-cmd --permanent --zone=public --add-service=postgresql
[all servers]# firewall-cmd --permanent --zone=public --add-port=9999/tcp --add-port=9898/tcp --add-port=9000/tcp  --add-port=9694/tcp
[all servers]# firewall-cmd --permanent --zone=public --add-port=9999/udp --add-port=9898/udp --add-port=9000/udp  --add-port=9694/udp
[all servers]# firewall-cmd --reload

7.3.5. Pgpool-II Configuration

7.3.5.1. Common Settings

Here are the common settings on server1 , server2 and server3 .

When installing Pgpool-II from RPM, all the Pgpool-II configuration files are in /etc/pgpool-II . In this example, we copy the sample configuration file for streaming replicaton mode.

# cp /etc/pgpool-II/pgpool.conf.sample-stream /etc/pgpool-II/pgpool.conf

To allow Pgpool-II to accept all incoming connections, we set listen_addresses = '*' .

listen_addresses = '*'

Specifiy replication delay check user and password. In this example, we leave sr_check_user empty, and create the entry in pool_passwd . From Pgpool-II 4.0, if these parameters are left blank, Pgpool-II will first try to get the password for that specific user from sr_check_password file before using the empty password.

sr_check_user = 'pgpool'
sr_check_password = ''

Enable health check so that pgpool-II performs failover. Also, if the network is unstable, the health check fails even though the backend is running properly, failover or degenerate operation may occur. In order to prevent such incorrect detection of health check, we set health_check_max_retries = 3 . Specify health_check_user and health_check_password in the same way like sr_check_user and sr_check_password .

health_check_period = 5
                                   # Health check period
                                   # Disabled (0) by default
health_check_timeout = 30
                                   # Health check timeout
                                   # 0 means no timeout
health_check_user = 'pgpool'
health_check_password = ''

health_check_max_retries = 3

Specify the PostgreSQL backend informations. Multiple backends can be specified by adding a number at the end of the parameter name.

# - Backend Connection Settings -

backend_hostname0 = 'server1'
                                   # Host name or IP address to connect to for backend 0
backend_port0 = 5432
                                   # Port number for backend 0
backend_weight0 = 1
                                   # Weight for backend 0 (only in load balancing mode)
backend_data_directory0 = '/var/lib/pgsql/11/data'
                                   # Data directory for backend 0
backend_flag0 = 'ALLOW_TO_FAILOVER'
                                   # Controls various backend behavior
                                   # ALLOW_TO_FAILOVER or DISALLOW_TO_FAILOVER
backend_hostname1 = 'server2'
backend_port1 = 5432
backend_weight1 = 1
backend_data_directory1 = '/var/lib/pgsql/11/data'
backend_flag1 = 'ALLOW_TO_FAILOVER'

backend_hostname2 = 'server3'
backend_port2 = 5432
backend_weight2 = 1
backend_data_directory2 = '/var/lib/pgsql/11/data'
backend_flag2 = 'ALLOW_TO_FAILOVER'

7.3.5.2. Failover configuration

Specify failover.sh script to be executed after failover in failover_command parameter. If we use 3 PostgreSQL servers, we need to specify follow_master_command to run after failover on the primary node failover. In case of two PostgreSQL servers, follow_master_command setting is not necessary.

failover_command = '/etc/pgpool-II/failover.sh %d %h %p %D %m %H %M %P %r %R'
follow_master_command = '/etc/pgpool-II/follow_master.sh %d %h %p %D %m %M %H %P %r %R'

Create /etc/pgpool-II/failover.sh , and add execute permission.

# vi /etc/pgpool-II/failover.sh
# vi /etc/pgpool-II/follow_master.sh
# chmod +x /etc/pgpool-II/{failover.sh,follow_master.sh}

/etc/pgpool-II/failover.sh

#!/bin/bash
# This script is run by failover_command.

set -o xtrace
exec > >(logger -i -p local1.info) 2>&1

# Special values:
#   %d = node id
#   %h = host name
#   %p = port number
#   %D = database cluster path
#   %m = new master node id
#   %H = hostname of the new master node
#   %M = old master node id
#   %P = old primary node id
#   %r = new master port number
#   %R = new master database cluster path
#   %% = '%' character

FAILED_NODE_ID="$1"
FAILED_NODE_HOST="$2"
FAILED_NODE_PORT="$3"
FAILED_NODE_PGDATA="$4"
NEW_MASTER_NODE_ID="$5"
NEW_MASTER_NODE_HOST="$6"
OLD_MASTER_NODE_ID="$7"
OLD_PRIMARY_NODE_ID="$8"
NEW_MASTER_NODE_PORT="$9"
NEW_MASTER_NODE_PGDATA="${10}"

PGHOME=/usr/pgsql-11

logger -i -p local1.info failover.sh: start: failed_node_id=$FAILED_NODE_ID old_primary_node_id=$OLD_PRIMARY_NODE_ID \
	failed_host=$FAILED_NODE_HOST new_master_host=$NEW_MASTER_NODE_HOST

# If standby node is down, skip failover.
if [ $FAILED_NODE_ID -ne $OLD_PRIMARY_NODE_ID ]; then
    logger -i -p local1.info failover.sh: Standby node is down. Skipping failover.
    exit 0
fi

# Promote standby node.
logger -i -p local1.info failover.sh: ssh: postgres@$NEW_MASTER_NODE_HOST pg_ctl promote
if [ $UID -eq 0 ]; then
    su postgres -c "ssh -T -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null \
    postgres@$NEW_MASTER_NODE_HOST ${PGHOME}/bin/pg_ctl -D ${NEW_MASTER_NODE_PGDATA} -w promote"
else
    ssh -T -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null \
    postgres@$NEW_MASTER_NODE_HOST ${PGHOME}/bin/pg_ctl -D ${NEW_MASTER_NODE_PGDATA} -w promote
fi

if [[ $? -ne 0 ]]; then
    logger -i -p local1.error failover.sh: new_master_host=$NEW_MASTER_NODE_HOST promote failed
    exit 1
fi

logger -i -p local1.info failover.sh: end: new_master_node_id=$NEW_MASTER_NODE_ID started as the primary node
exit 0

/etc/pgpool-II/follow_master.sh

#!/bin/bash
# This script is run after failover_command to recover the slave from the new primary.

set -o xtrace
exec > >(logger -i -p local1.info) 2>&1

# special values:  %d = node id
#                  %h = host name
#                  %p = port number
#                  %D = database cluster path
#                  %m = new master node id
#                  %M = old master node id
#                  %H = new master node host name
#                  %P = old primary node id
#                  %R = new master database cluster path
#                  %r = new master port number
#                  %% = '%' character
FAILED_NODE_ID="$1"
FAILED_NODE_HOST="$2"
FAILED_NODE_PORT="$3"
FAILED_NODE_PGDATA="$4"
NEW_MASTER_NODE_ID="$5"
OLD_MASTER_NODE_ID="$6"
NEW_MASTER_NODE_HOST="$7"
OLD_PRIMARY_NODE_ID="$8"
NEW_MASTER_NODE_PORT="$9"
NEW_MASTER_NODE_PGDATA="${10}"

PGHOME=/usr/pgsql-11
ARCHIVEDIR=/var/lib/pgsql/archivedir
REPL_USER=repl
PCP_USER=pgpool
PGPOOL_PATH=/usr/bin
PCP_PORT=9898


# Recovery the slave from the new primary
logger -i -p local1.info follow_master.sh: start: pg_basebackup for $FAILED_NODE_ID

# Check the status of standby
ssh -T -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null \
    postgres@${FAILED_NODE_HOST} ${PGHOME}/bin/pg_ctl -w -D ${FAILED_NODE_PGDATA} status >/dev/null 2>&1

# If slave is running, recover the slave from the new primary.
if [[ $? -eq 0 ]]; then

    # Execute pg_basebackup at slave
    ssh -T -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null postgres@${FAILED_NODE_HOST} "
        ${PGHOME}/bin/pg_ctl -w -m f -D ${FAILED_NODE_PGDATA} stop

        rm -rf ${FAILED_NODE_PGDATA}
        ${PGHOME}/bin/pg_basebackup -h ${NEW_MASTER_NODE_HOST} -U ${REPL_USER} -p ${NEW_MASTER_NODE_PORT} -D ${FAILED_NODE_PGDATA} -X stream -R

        if [[ $? -ne 0 ]]; then
            logger -i -p local1.error follow_master.sh: end: pg_basebackup failed
            exit 1
        fi
        rm -rf ${ARCHIVEDIR}/*
cat >> ${FAILED_NODE_PGDATA}/recovery.conf << EOT
restore_command = 'scp ${NEW_MASTER_NODE_HOST}:${ARCHIVEDIR}/%f %p'
EOT
        $PGHOME/bin/pg_ctl -l /dev/null -w -D ${FAILED_NODE_PGDATA} start
    "

    if [[ $? -eq 0 ]]; then

        # Run pcp_attact_node to attach this slave to Pgpool-II.
        ${PGPOOL_PATH}/pcp_attach_node -w -h localhost -U ${PCP_USER} -p ${PCP_PORT} -n ${FAILED_NODE_ID}

        if [[ $? -ne 0 ]]; then
            logger -i -p local1.error follow_master.sh: end: pcp_attach_node failed
            exit 1
        else
            logger -i -p local1.error follow_master.sh: end: follow master failed
            exit 1
        fi

else
    logger -i -p local1.info follow_master.sh: failed_nod_id=${FAILED_NODE_ID} is not running. skipping follow master command.
    exit 0
fi

logger -i -p local1.info follow_master.sh: end: follow master is finished
exit 0

7.3.5.3. Pgpool-II Online Recovery Configurations

Next, in order to perform online recovery with Pgpool-II we specify the PostgreSQL user name and online recovery command recovery_1st_stage . Because Supergroup privilege of PostgreSQL is required for online recovery, we specify postgres user to recovery_user . Then, we create recovery_1st_stage and pgpool_remote_start in database cluster directory of PostgreSQL primary server (server1), and add execute permission.

recovery_user = 'postgres'
                                   # Online recovery user
recovery_password = ''
                                   # Online recovery password

recovery_1st_stage_command = 'recovery_1st_stage'

[server1]# su - postgres
[server1]$ vi /var/lib/pgsql/11/data/recovery_1st_stage
[server1]$ vi /var/lib/pgsql/11/data/pgpool_remote_start
[server1]$ chmod +x /var/lib/pgsql/11/data/{recovery_1st_stage,pgpool_remote_start}

/var/lib/pgsql/11/data/recovery_1st_stage

#!/bin/bash
# This script is run by recovery_1st_stage to recovery the slave from the primary.

set -o xtrace
exec > >(logger -i -p local1.info) 2>&1

PRIMARY_NODE_PGDATA="$1"
DEST_NODE_HOST="$2"
DEST_NODE_PGDATA="$3"
PRIMARY_NODE_PGPORT=$4

PRIMARY_NODE_HOST=$(hostname -s)
PGHOME=/usr/pgsql-11
ARCHIVEDIR=/var/lib/pgsql/archivedir
REPLUSER=repl


logger -i -p local1.info online_recovery.sh: start: pg_basebackup for $DEST_NODE_HOST

# Run pg_basebackup to recovery the slave from the primary
ssh -T -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null postgres@$DEST_NODE_HOST "
rm -rf $DEST_NODE_PGDATA
${PGHOME}/bin/pg_basebackup -h $PRIMARY_NODE_HOST -U $REPLUSER -p $PRIMARY_NODE_PGPORT -D $DEST_NODE_PGDATA -X stream -R
"
if [[ $? -ne 0 ]]; then
    logger -i -p local1.error online_recovery.sh: end: pg_basebackup failed. online recovery failed.
    exit 1
fi

ssh -T -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null postgres@$DEST_NODE_HOST "
rm -rf $ARCHIVEDIR/*
cat >> $DEST_NODE_PGDATA/recovery.conf << EOT
restore_command = 'scp $PRIMARY_NODE_HOST:$archivedir/%f %p'
EOT
"

if [[ $? -ne 0 ]]; then
    logger -i -p local1.error online_recovery.sh: end: online recovery failed
    exit 1
else
    logger -i -p local1.info online_recovery.sh: end: online recovery is finished
    exit 0
fi

/var/lib/pgsql/11/data/pgpool_remote_start

#!/bin/bash
# This script is run to start slave node after recovery.

set -o xtrace
exec > >(logger -i -p local1.info) 2>&1

PGHOME=/usr/pgsql-11
DEST_HOST="$1"
DEST_HOST_PGDATA="$2"


logger -i -p local1.info pgpool_remote_start: start: remote start PostgreSQL@$DEST_HOST

# Start slave node
ssh -T -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null postgres@$DEST_HOST $PGHOME/bin/pg_ctl -l /dev/null -w -D $DEST_HOST_PGDATA start

if [[ $? -ne 0 ]]; then
    logger -i -p local1.info  pgpool_remote_start: $DEST_HOST start failed.
    exit 1
fi

logger -i -p local1.info pgpool_remote_start: end: $DEST_HOST PostgreSQL started successfully.

In order to use the online recovery functionality, the functions of pgpool_recovery , pgpool_remote_start , pgpool_switch_xlog are required, so we need install pgpool_recovery on template1 of PostgreSQL server server1 .

[server1]# su - postgres
[server1]$ psql template1 -c "CREATE EXTENSION pgpool_recovery"

7.3.5.4. Client Authentication Configuration

Because in the section Before Starting , we already set PostgreSQL authentication method to scram-sha-256 , it is necessary to set a client authentication by Pgpool-II to connect to backend nodes. Please note that only AES encrypted password or clear text passowrd can be specified in health_check_password , sr_check_password , wd_lifecheck_password , recovery_password in pgpool.conf . When installing with RPM, the Pgpool-II configuration file pool_hba.conf is in /etc/pgpool-II . By default, pool_hba authentication is disabled, and set enable_pool_hba = on to enable it.

enable_pool_hba = on

The format of pool_hba.conf file follows very closely PostgreSQL's pg_hba.conf format. Set pgpool and postgres user's authentication method to scram-sha-256 .

host    all         pgpool           0.0.0.0/0          scram-sha-256
host    all         postgres         0.0.0.0/0          scram-sha-256

The default password file name for authentication is pool_passwd . To use scram-sha-256 authentication, the decryption key to decrypt the passwords is required. We create the .pgpoolkey file in root user's home directory.

[all servers]# echo 'some string' > ~/.pgpoolkey 
[all servers]# chmod 600 ~/.pgpoolkey

Execute command pg_enc -m -k /path/to/.pgpoolkey -u username -p to regist user name and AES encrypted password in file pool_passwd . If pool_passwd doesn't exist yet, it will be created in the same directory as pgpool.conf .

[all servers]# pg_enc -m -k /root/.pgpoolkey -u pgpool -p
db password: [pgpool user's password]
[all servers]# pg_enc -m -k /root/.pgpoolkey -u postgres -p
db password: [postgres user's passowrd]

# cat /etc/pgpool-II/pool_passwd 
pgpool:AESheq2ZMZjynddMWk5sKP/Rw==
postgres:AESHs/pWL5rtXy2IwuzroHfqg==

7.3.5.5. Watchdog Configuration

Enable watchdog functionality on server1 , server2 , server3 .

use_watchdog = on

Specify virtual IP address that accepts connections from clients on server1 , server2 , server3 . Ensure that the IP address set to virtual IP isn't used yet.

delegate_IP = '192.168.137.150'

To bring up/down the virtual IP and send the ARP requests, we set if_up_cmd , if_down_cmd and arping_cmd . The network interface used in this example is "enp0s8".

if_up_cmd = 'ip addr add $_IP_$/24 dev enp0s8 label enp0s8:0'
                                    # startup delegate IP command
if_down_cmd = 'ip addr del $_IP_$/24 dev enp0s8'
                                    # shutdown delegate IP command
arping_cmd = 'arping -U $_IP_$ -w 1 -I enp0s8'
                                    # arping command

Set if_cmd_path and arping_path according to the command path.

if_cmd_path = '/sbin'
                                    # path to the directory where if_up/down_cmd exists
arping_path = '/usr/sbin'
                                    # arping command path

Specify the hostname and port number of each Pgpool-II server.

server1

wd_hostname = 'server1'
wd_port = 9000

server2

wd_hostname = 'server2'
wd_port = 9000

server3

wd_hostname = 'server3'
wd_port = 9000

Specify the hostname, Pgpool-II port number, and watchdog port number of monitored Pgpool-II servers on each Pgpool-II server.

server1

# - Other pgpool Connection Settings -

other_pgpool_hostname0 = 'server2'
                                    # Host name or IP address to connect to for other pgpool 0
                                    # (change requires restart)
other_pgpool_port0 = 9999
                                    # Port number for other pgpool 0
                                    # (change requires restart)
other_wd_port0 = 9000
                                    # Port number for other watchdog 0
                                    # (change requires restart)
other_pgpool_hostname1 = 'server3'
other_pgpool_port1 = 9999
other_wd_port1 = 9000

server2

# - Other pgpool Connection Settings -

other_pgpool_hostname0 = 'server1'
                                    # Host name or IP address to connect to for other pgpool 0
                                    # (change requires restart)
other_pgpool_port0 = 9999
                                    # Port number for other pgpool 0
                                    # (change requires restart)
other_wd_port0 = 9000
                                    # Port number for other watchdog 0
                                    # (change requires restart)
other_pgpool_hostname1 = 'server3'
other_pgpool_port1 = 9999
other_wd_port1 = 9000

server3

# - Other pgpool Connection Settings -

other_pgpool_hostname0 = 'server1'
                                    # Host name or IP address to connect to for other pgpool 0
                                    # (change requires restart)
other_pgpool_port0 = 9999
                                    # Port number for other pgpool 0
                                    # (change requires restart)
other_wd_port0 = 9000
                                    # Port number for other watchdog 0
                                    # (change requires restart)
other_pgpool_hostname1 = 'server2'
other_pgpool_port1 = 9999
other_wd_port1 = 9000

Specify the hostname and port number of destination for sending heartbeat signal on server1 , server2 , server3 .

server1

heartbeat_destination0 = 'server2'
                                    # Host name or IP address of destination 0
                                    # for sending heartbeat signal.
                                    # (change requires restart)
heartbeat_destination_port0 = 9694
                                    # Port number of destination 0 for sending
                                    # heartbeat signal. Usually this is the
                                    # same as wd_heartbeat_port.
                                    # (change requires restart)
heartbeat_device0 = ''
                                    # Name of NIC device (such like 'eth0')
                                    # used for sending/receiving heartbeat
                                    # signal to/from destination 0.
                                    # This works only when this is not empty
                                    # and pgpool has root privilege.
                                    # (change requires restart)

heartbeat_destination1 = 'server3'
heartbeat_destination_port1 = 9694
heartbeat_device1 = ''

server2

heartbeat_destination0 = 'server1'
                                    # Host name or IP address of destination 0
                                    # for sending heartbeat signal.
                                    # (change requires restart)
heartbeat_destination_port0 = 9694
                                    # Port number of destination 0 for sending
                                    # heartbeat signal. Usually this is the
                                    # same as wd_heartbeat_port.
                                    # (change requires restart)
heartbeat_device0 = ''
                                    # Name of NIC device (such like 'eth0')
                                    # used for sending/receiving heartbeat
                                    # signal to/from destination 0.
                                    # This works only when this is not empty
                                    # and pgpool has root privilege.
                                    # (change requires restart)

heartbeat_destination1 = 'server3'
heartbeat_destination_port1 = 9694
heartbeat_device1 = ''

server3

heartbeat_destination0 = 'server1'
                                    # Host name or IP address of destination 0
                                    # for sending heartbeat signal.
                                    # (change requires restart)
heartbeat_destination_port0 = 9694
                                    # Port number of destination 0 for sending
                                    # heartbeat signal. Usually this is the
                                    # same as wd_heartbeat_port.
                                    # (change requires restart)
heartbeat_device0 = ''
                                    # Name of NIC device (such like 'eth0')
                                    # used for sending/receiving heartbeat
                                    # signal to/from destination 0.
                                    # This works only when this is not empty
                                    # and pgpool has root privilege.
                                    # (change requires restart)

heartbeat_destination1 = 'server2'
heartbeat_destination_port1 = 9694
heartbeat_device1 = ''

7.3.5.6. /etc/sysconfig/pgpool Configuration

If you want to ignore the pgpool_status file at startup of Pgpool-II , add "- D" to the start option OPTS to /etc/sysconfig/pgpool .

[all servers]# vi /etc/sysconfig/pgpool 
...
OPTS=" -D -n"

7.3.5.7. Logging

In the example, we output Pgpool-II 's log to syslog .

log_destination = 'syslog'
                                   # Where to log
                                   # Valid values are combinations of stderr,
                                   # and syslog. Default to stderr.

syslog_facility = 'LOCAL1'
                                   # Syslog local facility. Default to LOCAL0

Create Pgpool-II log file.

[all servers]# mkdir /var/log/pgpool-II
[all servers]# touch /var/log/pgpool-II/pgpool.log

Edit config file of syslog /etc/rsyslog.conf .

[all servers]# vi /etc/rsyslog.conf
...
*.info;mail.none;authpriv.none;cron.none;LOCAL1.none    /var/log/messages
LOCAL1.*                                                /var/log/pgpool-II/pgpool.log

Setting logrotate same as /var/log/messages .

[all servers]# vi /etc/logrotate.d/syslog
...
/var/log/messages
/var/log/pgpool-II/pgpool.log
/var/log/secure

Restart rsyslog service.

[all servers]# systemctl restart rsyslog

7.3.5.8. PCP Command Configuration

Since user authentication is required to use the PCP command, specify user name and md5 encrypted password in pcp.conf . Here we create the encrypted password for pgpool user, and add " username:encrypted password " in /etc/pgpool-II/pcp.conf .

[all servers]# echo 'pgpool:'`pg_md5 PCP passowrd` >> /etc/pgpool-II/pcp.conf

7.3.5.9. .pcppass

Since follow_master_command script has to execute PCP command without entering the password, we create .pcppass in the home directory of Pgpool-II startup user (root user).

[all servers]# echo 'localhost:9898:pgpool:pgpool' > ~/.pcppass
[all servers]# chmod 600 ~/.pcppass

The settings of Pgpool-II is completed.

7.3.6. Starting/Stopping Pgpool-II

Next we start Pgpool-II . Before starting Pgpool-II , please start PostgreSQL servers first. Also, when stopping PostgreSQL , it is necessary to stop Pgpool-II first.

Starting Pgpool-II

In section Before Starting , we already set the auto-start of Pgpool-II . To start Pgpool-II , restart the whole system or execute the following command.
```
# systemctl start pgpool.service
          
```

Stopping Pgpool-II

# systemctl stop pgpool.service

7.3.7. How to use

Let's start to use Pgpool-II . First, let's start Pgpool-II on server1 , server2 , server3 by using the following command.

# systemctl start pgpool.service

7.3.7.1. Set up PostgreSQL standby server

First, we should set up PostgreSQL standby server by using Pgpool-II online recovery functionality. Ensure that recovery_1st_stage and pgpool_remote_start scripts used by pcp_recovery_node command are in database cluster directory of PostgreSQL primary server ( server1 ).

# pcp_recovery_node -h 192.168.137.150 -p 9898 -U pgpool -n 1
Password: 
pcp_recovery_node -- Command Successful

# pcp_recovery_node -h 192.168.137.150 -p 9898 -U pgpool -n 2
Password: 
pcp_recovery_node -- Command Successful

After executing pcp_recovery_node command, vertify that server2 and server3 are started as PostgreSQL standby server.

# psql -h 192.168.137.150 -p 9999 -U pgpool postgres -c "show pool_nodes"
 node_id | hostname | port | status | lb_weight |  role   | select_cnt | load_balance_node | replication_delay | last_status_change  
---------+----------+------+--------+-----------+---------+------------+-------------------+-------------------+---------------------
 0       | server1  | 5432 | up     | 0.333333  | primary | 0          | false             | 0                 | 2019-02-18 11:26:31
 1       | server2  | 5432 | up     | 0.333333  | standby | 0          | true              | 0                 | 2019-02-18 11:27:49
 2       | server3  | 5432 | up     | 0.333333  | standby | 0          | false             | 0                 | 2019-02-18 11:27:49

7.3.7.2. Switching active/standby watchdog

Confirm the watchdog status by using pcp_watchdog_info . The Pgpool-II server which is started first run as MASTER .

# pcp_watchdog_info -h 192.168.137.150 -p 9898 -U pgpool
Password: 
3 YES server1:9999 Linux server1 server1

server1:9999 Linux server1 server1 9999 9000 4 MASTER  #The Pgpool-II server started first becames "MASTER".
server2:9999 Linux server2 server2 9999 9000 7 STANDBY #run as standby
server3:9999 Linux server3 server3 9999 9000 7 STANDBY #run as standby

Stop active server server1 , then server2 or server3 will be promoted to active server. To stop server1 , we can stop Pgpool-II service or shutdown the whole system. Here, we stop Pgpool-II service.

[server1]# systemctl stop pgpool.service

# pcp_watchdog_info -p 9898 -h 192.168.137.150 -U pgpool
Password: 
3 YES server2:9999 Linux server2 server2

server2:9999 Linux server2 server2 9999 9000 4 MASTER     #server2 is promoted to MASTER
server1:9999 Linux server1 server1 9999 9000 10 SHUTDOWN  #server1 is stopped
server3:9999 Linux server3 server3 9999 9000 7 STANDBY    #server3 runs as STANDBY

Start Pgpool-II ( server1 ) which we have stopped again, and vertify that server1 runs as a standby.

[server1]# systemctl start pgpool.service

[server1]# pcp_watchdog_info -p 9898 -h 192.168.137.150 -U pgpool
Password: 
3 YES server2:9999 Linux server2 server2

server2:9999 Linux server2 server2 9999 9000 4 MASTER
server1:9999 Linux server1 server1 9999 9000 7 STANDBY
server3:9999 Linux server3 server3 9999 9000 7 STANDBY

7.3.7.3. Failover

First, use psql to connect to PostgreSQL via virtual IP, and verify the backend informations.

# psql -h 192.168.137.150 -p 9999 -U pgpool postgres -c "show pool_nodes"
 node_id | hostname | port | status | lb_weight |  role   | select_cnt | load_balance_node | replication_delay | last_status_change  
---------+----------+------+--------+-----------+---------+------------+-------------------+-------------------+---------------------
 0       | server1  | 5432 | up     | 0.333333  | primary | 0          | false             | 0                 | 2019-02-18 13:08:02
 1       | server2  | 5432 | up     | 0.333333  | standby | 0          | false             | 0                 | 2019-02-18 13:21:56
 2       | server3  | 5432 | up     | 0.333333  | standby | 0          | true              | 0                 | 2019-02-18 13:21:56

Next, stop primary PostgreSQL server server1 , and verify automatic failover.

[server1]$ pg_ctl -D /var/lib/pgsql/11/data -m immediate stop

After stopping PostgreSQL on server1 , failover occurs and PostgreSQL on server2 becomes new primary DB.

# psql -h 192.168.137.150 -p 9999 -U pgpool postgres -c "show pool_nodes"
 node_id | hostname | port | status | lb_weight |  role   | select_cnt | load_balance_node | replication_delay | last_status_change  
---------+----------+------+--------+-----------+---------+------------+-------------------+-------------------+---------------------
 0       | server1  | 5432 | down   | 0.333333  | standby | 0          | false             | 0                 | 2019-02-18 13:22:25
 1       | server2  | 5432 | up     | 0.333333  | primary | 0          | true              | 0                 | 2019-02-18 13:22:25
 2       | server3  | 5432 | up     | 0.333333  | standby | 0          | false             | 0                 | 2019-02-18 13:22:28

server3 is running as standby of new primary server2 .

[server3]# psql -h server3 -p 5432 -U pgpool postgres -c "select pg_is_in_recovery()"
 pg_is_in_recovery 
-------------------
 t

[server2]# su - postgres
$ psql
postgres=# select pg_is_in_recovery();
 pg_is_in_recovery 
-------------------
 f

postgres=# select * from pg_stat_replication;
-[ RECORD 1 ]----+------------------------------
pid              | 11915
usesysid         | 16385
usename          | repl
application_name | walreceiver
client_addr      | 192.168.137.103
client_hostname  | 
client_port      | 37834
backend_start    | 2019-02-18 13:22:27.472038+09
backend_xmin     | 
state            | streaming
sent_lsn         | 0/8E000060
write_lsn        | 0/8E000060
flush_lsn        | 0/8E000060
replay_lsn       | 0/8E000060
write_lag        | 
flush_lag        | 
replay_lag       | 
sync_priority    | 0
sync_state       | async

7.3.7.4. Online Recovery

Here, we use Pgpool-II online recovery functionality to restore server1 (old primary server) as a standby. Before restoring the old primary server, please ensure that recovery_1st_stage and pgpool_remote_start scripts exist in database cluster directory of current primary server server2 .

# pcp_recovery_node -h 192.168.137.150 -p 9898 -U pgpool -n 0
Password: 
pcp_recovery_node -- Command Successful

Then verify that server1 is started as a standby.

# psql -h 192.168.137.150 -p 9999 -U pgpool postgres -c "show pool_nodes"
 node_id | hostname | port | status | lb_weight |  role   | select_cnt | load_balance_node | replication_delay | last_status_change  
---------+----------+------+--------+-----------+---------+------------+-------------------+-------------------+---------------------
 0       | server1  | 5432 | up     | 0.333333  | standby | 0          | false             | 0                 | 2019-02-18 13:27:44
 1       | server2  | 5432 | up     | 0.333333  | primary | 0          | false             | 0                 | 2019-02-18 13:22:25
 2       | server3  | 5432 | up     | 0.333333  | standby | 0          | true              | 0                 | 2019-02-18 13:22:28