pg13 交换主备库的角色（failover）

数据库版本信息：

postgres=# select version();
                                                 version
---------------------------------------------------------------------------------------------------------
 PostgreSQL 13.4 on x86_64-pc-linux-gnu, compiled by gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-39), 64-bit
(1 row)

主备数据库信息：

主：127.0.0.1 5433 目录  /data/s001

备：127.0.0.1 5432  目录/data/master

01.模拟主机故障，停止主数据库

-bash-4.2$ pg_ctl -D /data/s001/ stop
waiting for server to shut down....2021-10-24 16:55:32.054 CST [10171] LOG:  received fast shutdown request
2021-10-24 16:55:32.064 CST [10171] LOG:  aborting any active transactions
2021-10-24 16:55:32.071 CST [10171] LOG:  background worker "logical replication launcher" (PID 10178) exited with exit code 1
2021-10-24 16:55:32.071 CST [10173] LOG:  shutting down
2021-10-24 16:55:32.132 CST [10611] LOG:  replication terminated by primary server
2021-10-24 16:55:32.132 CST [10611] DETAIL:  End of WAL reached on timeline 1 at 0/3011FF8.
2021-10-24 16:55:32.132 CST [10611] FATAL:  could not send end-of-streaming message to primary: no COPY in progress
2021-10-24 16:55:32.134 CST [10607] LOG:  record length 1207964928 at 0/3011FF8 too long
2021-10-24 16:55:32.149 CST [10171] LOG:  database system is shut down
2021-10-24 16:55:32.153 CST [14378] FATAL:  could not connect to the primary server: server closed the connection unexpectedly
                This probably means the server terminated abnormally
                before or while processing the request.
 done
server stopped

02，激活备库

-bash-4.2$ pg_ctl -D /data/master promote
waiting for server to promote.... done
server promoted

#备库提升主库日志：
2021-10-24 16:56:36.020 CST [10607] LOG:  received promote request
2021-10-24 16:56:36.020 CST [10607] LOG:  redo done at 0/3011F80
2021-10-24 16:56:36.020 CST [10607] LOG:  last completed transaction was at log time 2021-10-24 16:49:03.893587+08
2021-10-24 16:56:36.044 CST [10607] LOG:  selected new timeline ID: 2
2021-10-24 16:56:36.244 CST [10607] LOG:  archive recovery complete
2021-10-24 16:56:36.269 CST [10606] LOG:  database system is ready to accept connections

检查数据库状态，已经提升为主数据库：

postgres=# select pg_is_in_recovery();
 pg_is_in_recovery
-------------------
 f
(1 row)

03，原主库，重新提升为备机

重新同步数据库：

walpg_rewind -D $PGDATA --source-server='host=10.0.3.102 user=postgres password=’

或者使用

pg_basebackup