postgres主从切换后，原主重新加入集群报错:pg_rewind: error: could not open file

一、用keepalived进行脚本主从切换，当原主重新加入集群报错

[root@hlcc_slave data]# /usr/pgsql-13/bin/pg_rewind -D '/home/postgres/data' --source-server='host=172.28.17.142 port=5876 user=repmgr password=hl95repmgr dbname=repmgr connect_timeout=2'
pg_rewind: error: cannot be executed by "root"
You must run pg_rewind as the PostgreSQL superuser.
[root@hlcc_slave data]# su - postgres
Last login: Wed Aug  3 09:31:30 CST 2022 on pts/0
[postgres@hlcc_slave ~]$ /usr/pgsql-13/bin/pg_rewind -D '/home/postgres/data' --source-server='host=172.28.17.142 port=5876 user=repmgr password=hl95repmgr dbname=repmgr connect_timeout=2'
pg_rewind: servers diverged at WAL location 0/140000A0 on timeline 9
pg_rewind: error: could not open file "/home/postgres/data/pg_wal/000000090000000000000014": 

二、查看master以及slave节点数据库的data/pg_wal，均没有这个文件

（如果有，可以直接复制，再重新加入集群即可）

 

三、重新删除data目录，重新克隆主库，并注册从库

1、删除data目录

su - postgres
mv data data.yyyymmdd
mkdir data

2、克隆新主库

[postgres@hlcc_slave ~]$ repmgr -h hlcc_master -p 5876 -U repmgr -d repmgr -f /etc/repmgr/13/repmgr.conf standby clone -F
NOTICE: destination directory "/home/postgres/data" provided
INFO: connecting to source node
DETAIL: connection string is: host=hlcc_master port=5876 user=repmgr dbname=repmgr
DETAIL: current installation size is 158 MB
NOTICE: checking for available walsenders on the source node (2 required)
NOTICE: checking replication connections can be made to the source server (2 required)
INFO: checking and correcting permissions on existing directory "/home/postgres/data"
NOTICE: starting backup (using pg_basebackup)...
HINT: this may take some time; consider using the -c/--fast-checkpoint option
INFO: executing:
  /usr/pgsql-13/bin/pg_basebackup -l "repmgr base backup"  -D /home/postgres/data -h hlcc_master -p 5876 -U repmgr -X stream -S repmgr_slot_2 
NOTICE: standby clone (using pg_basebackup) complete
NOTICE: you can now start your PostgreSQL server
HINT: for example: /home/postgres/bin/postgresql  start
HINT: after starting the server, you need to re-register this standby with "repmgr standby register --force" to update the existing node record

3、启动当前数据库

[postgres@hlcc_slave home]$ postgresql start
Starting PostgreSQL: OK
[postgres@hlcc_slave home]$ postgresql status
0

4、注册为从库

[postgres@hlcc_slave home]$ repmgr standby register -f /etc/repmgr/13/repmgr.conf --force
INFO: connecting to local node "hlcc_slave" (ID: 2)
INFO: connecting to primary database
INFO: standby registration complete
NOTICE: standby node "hlcc_slave" (ID: 2) successfully registered

5、查看主从状态

[postgres@hlcc_slave home]$ repmgr -f /etc/repmgr/13/repmgr.conf   cluster show
 ID | Name        | Role    | Status    | Upstream    | Location | Priority | Timeline | Connection string                                                                           
----+-------------+---------+-----------+-------------+----------+----------+----------+----------------------------------------------------------------------------------------------
 1  | hlcc_master | primary | * running |             | default  | 100      | 11       | host=172.28.17.142 port=5876 user=repmgr password=hl95repmgr dbname=repmgr connect_timeout=2
 2  | hlcc_slave  | standby |   running | hlcc_master | default  | 100      | 11       | host=172.28.17.143 port=5876 user=repmgr password=hl95repmgr dbname=repmgr connect_timeout=2
[postgres@hlcc_slave home]$