Postgresql

Pgpool

2018-10-23 19:00:00 +0800

Optional Solution

1. corosync & pacemaker & heartbeat script

   POC failed, centos 7.2 x64

2. pgpool II (middleware) & failover script & recovery script

   POC … processing

POC

Env

CentOS 7.2 x64

Installation

1. postgresql 9.6
2. pgpool II 96 v3.6.12

Version

9.6 & 3.6

Config

Hot-standby

1. backup command

   pg_basebackup -D .././. -F p -x -h primary_node -p 5432 -U postgres -P -v
   

Check pool node status

psql -h primary_node -p9999 -U postgres postgres

Pg_rewind command

su - postgres

pg_rewind --target-pgdata='/../../..' --source-server='host= port= dbname= user= password=' -P

note: the target server must be primary server, postgresql won’t let server to synchronize from standby server.

issue

1. watchdog reject issue, correct Other pgpool Connection Settings pgpool_port and wd_port, default is 9999 and 9000.

   This issue will lead watchdog process down, watchdog heartbeat signal will be rejected.

2. correct the rwx permission for root and postgres user.

3. recovery issue

4. pg_rewind command for base time line synchronize