HG_REPMGR autofailvoer自动故障转移

原文链接： https://support.highgo.com/#/index/docContent/fcf01ba201b12041

目录

文档用途

详细信息

文档用途

HG_REPMGR自动故障转移配置参考

详细信息

配置集群自动故障转移（failover），需要为集群中的每个节点开启 repmgrd 守护进程。当主节点出现故障后，会自动将合适的备节点提升为新主节点，继

续对外提供服务。示例如下。

 

1. 配置 postgresql.replication.conf 文件（所有节点）

在上述 postgresql.replication.conf 的基础上，添加如下参数： 

 

shared_preload_libraries = 'repmgr'   或者   alter system set shared_preload_libraries =pg_pathman,timescaledb,repmgr;

 

重启数据库：  

 

pg_ctl restart

 

 

2. 配置 hg_repmgr.conf（所有节点）

在现有的 hg_repmgr.conf 文件中添加如下参数： 

 

failover=automatic      promote_command='repmgr -f /opt/highgo/5.6.1/conf/hg_repmgr.conf standby promote' follow_command='repmgr -f /opt/highgo/5.6.1/conf/hg_repmgr.conf standby follow --upstream-node-id=%n'

 

 

如果需要将 repmgr 的日志定位到固定的日志文件可添加 log_file 参数，如 下：

 

log_file='/opt/highgo/5.6.1/conf/data/log/hg_repmgr.log'

  

为了防止上述日志文件不断膨胀，可配置系统的 logrotate。（详细步骤略）

 

3. 开启 repmgrd 进程（所有节点） 

 

repmgrd  -f /opt/highgo/5.6.1/conf/hg_repmgr.conf -d  -p /tmp/hg_repmgrd.pid   [highgo@dbrs conf]$ repmgrd  -d  -p /tmp/hg_repmgrd.pid [2019-05-06 14:02:42] [NOTICE] repmgrd (repmgrd 4.2) starting up [2019-05-06 14:02:42] [INFO] connecting to database "" [2019-05-06 14:02:43] [ERROR] repmgr extension not found on this node [2019-05-06 14:02:43] [DETAIL] repmgr extension is available but not installed in database "highgo" [2019-05-06 14:02:43] [HINT] check that this node is part of a repmgr cluster [highgo@dbrs conf]$   highgo=# \c You are now connected to database "highgo" as user "highgo".     create extension repmgr;   [highgo@dbrs conf]$ repmgrd  -f /opt/highgo/5.6.1/conf/hg_repmgr.conf -d  -p /tmp/hg_repmgrd.pid [2019-05-06 14:21:21] [NOTICE] repmgrd (repmgrd 4.2) starting up [2019-05-06 14:21:21] [INFO] connecting to database "host=dbrs user=hgrepmgr dbname=hgrepmgr connect_timeout=2" [highgo@dbrs conf]$ хϢ:  set_repmgrd_pid(): provided pidfile is /tmp/hg_repmgrd.pid [2019-05-06 14:21:21] [NOTICE] starting monitoring of node "dbrs" (ID: 1) [2019-05-06 14:21:21] [NOTICE] monitoring cluster primary "dbrs" (node ID: 1)   [highgo@dbrs2 conf]$ repmgrd  -f /opt/highgo/5.6.1/conf/hg_repmgr.conf -d  -p /tmp/hg_repmgrd.pid [2019-05-06 14:21:50] [NOTICE] repmgrd (repmgrd 4.2) starting up [2019-05-06 14:21:50] [INFO] connecting to database "host=dbrs2 user=hgrepmgr dbname=hgrepmgr connect_timeout=2" [highgo@dbrs2 conf]$ хϢ:  set_repmgrd_pid(): provided pidfile is /tmp/hg_repmgrd.pid [2019-05-06 14:21:50] [NOTICE] starting monitoring of node "dbrs2" (ID: 2) [2019-05-06 14:21:50] [INFO] monitoring connection to upstream node "dbrs" (node ID: 1)   [highgo@dbrs conf]$ ls -atl /tmp/hg_repmgrd.pid -rw-rw-r--. 1 highgo highgo 5 May  6 14:21 /tmp/hg_repmgrd.pid [highgo@dbrs conf]$ [highgo@dbrs2 conf]$ ls -atl /tmp/hg_repmgrd.pid -rw-rw-r--. 1 highgo highgo 5 May  6 14:21 /tmp/hg_repmgrd.pid [highgo@dbrs2 conf]$

 

 提示：这个后台进程，每次重启服务器，都要手动启动吗？

开发回复：目前是，后期会修改为自动

 

查看集群状态

 

[highgo@dbrs conf]$ repmgr -f /opt/highgo/5.6.1/conf/hg_repmgr.conf cluster show  ID | Name  | Role    | Status    | Upstream | Location | Connection string                                         ----+-------+---------+-----------+----------+----------+------------------------------------------------------------  1  | dbrs  | primary | * running |          | default  | host=dbrs user=hgrepmgr dbname=hgrepmgr connect_timeout=2  2  | dbrs2 | standby |   running | dbrs     | default  | host=dbrs2 user=hgrepmgr dbname=hgrepmgr connect_timeout=2 [highgo@dbrs conf]$

 

 

模拟主节点故障

 

1）在 node1 上关闭数据库 pg_ctl stop 2）在 node2 上查看集群状态 [highgo@dbrs2 conf]$ repmgr -f /opt/highgo/5.6.1/conf/hg_repmgr.conf cluster show  ID | Name  | Role    | Status    | Upstream | Location | Connection string                                         ----+-------+---------+-----------+----------+----------+------------------------------------------------------------  1  | dbrs  | primary | - failed  |          | default  | host=dbrs user=hgrepmgr dbname=hgrepmgr connect_timeout=2  2  | dbrs2 | primary | * running |          | default  | host=dbrs2 user=hgrepmgr dbname=hgrepmgr connect_timeout=2   WARNING: following issues were detected   - unable to connect to node "dbrs" (ID: 1) [highgo@dbrs2 conf]$   此时 node2 已经提升为 primary     日志 [highgo@dbrs2 conf]$ [2019-05-06 14:24:14] [WARNING] unable to connect to upstream node "dbrs" (node ID: 1) [2019-05-06 14:24:14] [INFO] checking state of node 1, 1 of 6 attempts [2019-05-06 14:24:14] [INFO] sleeping 10 seconds until next reconnection attempt [2019-05-06 14:24:24] [INFO] checking state of node 1, 2 of 6 attempts [2019-05-06 14:24:24] [INFO] sleeping 10 seconds until next reconnection attempt [2019-05-06 14:24:34] [INFO] checking state of node 1, 3 of 6 attempts [2019-05-06 14:24:34] [INFO] sleeping 10 seconds until next reconnection attempt [2019-05-06 14:24:44] [INFO] checking state of node 1, 4 of 6 attempts [2019-05-06 14:24:44] [INFO] sleeping 10 seconds until next reconnection attempt [2019-05-06 14:24:54] [INFO] checking state of node 1, 5 of 6 attempts [2019-05-06 14:24:54] [INFO] sleeping 10 seconds until next reconnection attempt   [highgo@dbrs2 conf]$ [2019-05-06 14:25:04] [INFO] checking state of node 1, 6 of 6 attempts [2019-05-06 14:25:04] [WARNING] unable to reconnect to node 1 after 6 attempts [2019-05-06 14:25:04] [NOTICE] this node is the only available candidate and will now promote itself [2019-05-06 14:25:04] [INFO] promote_command is:   "repmgr -f /opt/highgo/5.6.1/conf/hg_repmgr.conf standby promote" NOTICE: promoting standby to primary DETAIL: promoting server "dbrs2" (ID: 2) using "/opt/highgo/5.6.1/bin/pg_ctl  -w -D '/opt/highgo/5.6.1/data' promote" DETAIL: waiting up to 60 seconds (parameter "promote_check_timeout") for promotion to complete NOTICE: STANDBY PROMOTE successful DETAIL: server "dbrs2" (ID: 2) was successfully promoted to primary [2019-05-06 14:25:10] [INFO] switching to primary monitoring mode [2019-05-06 14:25:10] [NOTICE] monitoring cluster primary "dbrs2" (node ID: 2)

更多详细信息请登录【瀚高技术支持平台】 查看https://support.highgo.com/#/index/docContent/fcf01ba201b12041