repmgr安裝使用

本文轉載自查看原文 2020-09-29 19:13 994 PostgreSQL

repmgr安裝使用

node1: 192.168.5.132
node2: 192.168.5.133

一、通用安裝，兩個節點都執行：
1.repmgr安裝：
Install the repository definition for your distribution and PostgreSQL version：
curl https://dl.2ndquadrant.com/default/release/get/11/rpm | sudo bash
sudo yum repolist

Install：
sudo yum -y install repmgr12

2.安裝PG12
yum localinstall postgresql12-*

二、主庫設置：
1.初始化數據庫：
cd /usr/pg-12/
initdb -D data

2.修改數據庫配置文件：
[postgres@node1 data]$ vim postgresql.conf
listen_addresses = '*'
shared_preload_libraries = 'repmgr'
wal_log_hints = on

可以添加：

synchronous_standby_names = 'ANY 1(*)'

[postgres@node1 data]$ vim pg_hba.conf
host all all 192.168.5.132/32 trust
host all all 192.168.5.133/32 trust

host replication all 192.168.5.132/32 trust
host replication all 192.168.5.133/32 trust

啟動數據庫，使上面參數生效。注意：在clone備庫的時候，沒法輸入密碼，因此用trust。

pg_ctl -D ./ start

3.修改配置文件：
vim /etc/repmgr/12/repmgr.conf

node_id=1
node_name=node1
conninfo='host=192.168.5.132 port=5432 user=postgres dbname=postgres'
data_directory='/usr/pgsql-12/data'

4.注冊主節點
[postgres@node1 data]$ repmgr primary register
INFO: connecting to primary database...
NOTICE: attempting to install extension "repmgr"
NOTICE: "repmgr" extension successfully installed
NOTICE: primary node record (ID: 1) registered

三、備庫設置
1.修改配置：
vim /etc/repmgr/12/repmgr.conf

node_id=2
node_name=node2
conninfo='host=192.168.5.133 port=5432 user=postgres dbname=postgres'
data_directory='/usr/pgsql-12/data'

2.clone備庫

[postgres@localhost pgsql-12]$ repmgr standby clone -h 192.168.5.132 -U postgres
NOTICE: destination directory "/usr/pgsql-12/data" provided
INFO: connecting to source node
DETAIL: connection string is: host=192.168.5.132 user=postgres
DETAIL: current installation size is 23 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)
WARNING: data checksums are not enabled and "wal_log_hints" is "off"
DETAIL: pg_rewind requires "wal_log_hints" to be enabled
INFO: creating directory "/usr/pgsql-12/data"...
NOTICE: starting backup (using pg_basebackup)...
HINT: this may take some time; consider using the -c/--fast-checkpoint option
INFO: executing:
  pg_basebackup -l "repmgr base backup"  -D /usr/pgsql-12/data -h 192.168.5.132 -p 5432 -U postgres -X stream 
NOTICE: standby clone (using pg_basebackup) complete
NOTICE: you can now start your PostgreSQL server
HINT: for example: pg_ctl -D /usr/pgsql-12/data start
HINT: after starting the server, you need to register this standby with "repmgr standby register"

3.啟動並注冊

[postgres@localhost pgsql-12]$ pg_ctl -D /usr/pgsql-12/data start
waiting for server to start....2020-03-08 10:43:32.861 CST [74044] LOG:  starting PostgreSQL 12.4 on x86_64-pc-linux-gnu, compiled by gcc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-23), 64-bit
2020-03-08 10:43:32.935 CST [74044] LOG:  listening on IPv4 address "0.0.0.0", port 5432
2020-03-08 10:43:32.968 CST [74044] LOG:  listening on IPv6 address "::", port 5432
2020-03-08 10:43:32.973 CST [74044] LOG:  listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
2020-03-08 10:43:33.004 CST [74044] LOG:  listening on Unix socket "/tmp/.s.PGSQL.5432"
2020-03-08 10:43:33.072 CST [74044] LOG:  redirecting log output to logging collector process
2020-03-08 10:43:33.072 CST [74044] HINT:  Future log output will appear in directory "log".
 done
server started
[postgres@localhost pgsql-12]$ repmgr standby register
INFO: connecting to local node "node2" (ID: 2)
INFO: connecting to primary database
WARNING: --upstream-node-id not supplied, assuming upstream node is primary (node ID 1)
INFO: standby registration complete
NOTICE: standby node "node2" (ID: 2) successfully registered

[postgres@localhost pgsql-12]$ repmgr cluster show
 ID | Name  | Role    | Status    | Upstream | Location | Priority | Timeline | Connection string                                          
----+-------+---------+-----------+----------+----------+----------+----------+-------------------------------------------------------------
 1  | node1 | primary | * running |          | default  | 100      | 1        | host=192.168.5.132 port=5432 user=postgres  dbname=postgres
 2  | node2 | standby |   running | node1    | default  | 100      | 1        | host=192.168.5.133 port=5432 user=postgres  dbname=postgres

failover='automatic'
promote_command='/usr/pgsql-12/bin/repmgr standby promote'
follow_command='/usr/pgsql-12/bin/repmgr standby follow'

failover參數有兩個
automatic：表示開啟故障自動切換
manual：不開啟故障自動切換

3.啟動repmgrd服務
[postgres@localhost pgsql-12]$ repmgrd -d
[2020-03-07 18:48:58] [NOTICE] repmgrd (repmgrd 5.1.0) starting up
[2020-03-07 18:48:58] [INFO] connecting to database "host=192.168.5.133 port=5432 user=postgres dbname=postgres"
[postgres@localhost pgsql-12]$ INFO: set_repmgrd_pid(): provided pidfile is /tmp/repmgrd.pid
[2020-03-07 18:48:58] [NOTICE] starting monitoring of node "node2" (ID: 2)
[2020-03-07 18:48:58] [INFO] "connection_check_type" set to "ping"
[2020-03-07 18:48:58] [INFO] monitoring connection to upstream node "node1" (ID: 1)

五、日志：
默認情況下，repmgr和repmgrd將日志輸出寫入 STDERR。
可以指定備用日志目標（文件或syslog）。

注意：
即使配置了另一個日志目標，repmgr應用程序本身也會繼續將日志輸出寫入到STDERR日志中，否則，由命令行操作產生的任何輸出都會“消失”在日志中。

在配置文件添加日志文件路徑：
[postgres@node1 repmgr]$ vim /etc/repmgr/12/repmgr.conf
log_file='/var/log/repmgr/repmgrd.log'

添加文件：
[root@node1 ~]# vim /etc/logrotate.d/repmgr

/var/log/repmgr/repmgrd.log {
        missingok
        compress
        rotate 52
        maxsize 100M
        weekly
        create 0600 postgres postgres
        postrotate
            /usr/bin/killall -HUP repmgrd
        endscript
}

六、添加witness節點
[postgres@localhost pgsql-12]$ repmgr witness register -h 192.168.5.133
INFO: connecting to witness node "node2" (ID: 2)
ERROR: provided node is a standby
HINT: a witness node must run on an independent primary server

七、命令使用
[postgres@localhost pgsql-12]$ repmgr --help
repmgr: replication management tool for PostgreSQL
Usage:
repmgr [OPTIONS] primary {register|unregister}
repmgr [OPTIONS] standby {register|unregister|clone|promote|follow|switchover}
repmgr [OPTIONS] node {status|check|rejoin|service}
repmgr [OPTIONS] cluster {show|event|matrix|crosscheck|cleanup}
repmgr [OPTIONS] witness {register|unregister}
repmgr [OPTIONS] service {status|pause|unpause}
repmgr [OPTIONS] daemon {start|stop}

3）停止repmgrd
配置repmgr.conf文件，添加命令：
repmgrd_service_start_command ='service repmgr-12 start'
repmgrd_service_stop_command ='service repmgr-12 stop'

執行停止：
[postgres@localhost pgsql-12]$ repmgr daemon stop
NOTICE: executing: "service repmgr-12 stop"
ERROR: repmgrd does not appear to have stopped after 15 seconds
HINT: use "repmgr service status" to confirm that repmgrd was successfully started
--但是：service repmgr-12 stop 沒有停止

換一個配置命令：
repmgrd_service_stop_command ='repmgr node service --list-actions --action=stop'

重新執行：
[postgres@localhost pgsql-12]$ repmgr daemon stop
NOTICE: executing: "repmgr node service --list-actions --action=stop"
ERROR: repmgrd does not appear to have stopped after 15 seconds
HINT: use "repmgr service status" to confirm that repmgrd was successfully started

仍然沒有停止掉

七、驗證
1）關閉repmgr進程，有什么影響？
repmgr不是進程，是一個插件，沒有可以停止的地方

2）關閉repmgrd進程，有什么影響？
只是repmgrd進程掉了，數據庫不會有影響。流復制集群正常~

3）如何重新加載repmgr.conf文件
殺掉進程，重新啟動：
[postgres@localhost pgsql-12]$ kill 74238
[postgres@localhost pgsql-12]$ [2020-03-07 19:44:15] [NOTICE] TERM signal received
[2020-03-07 19:44:15] [INFO] repmgrd terminating...

4）如果模擬切換？
手動停止主數據庫

主庫執行：

[postgres@node1 data]$ pg_ctl -D ./ stop
waiting for server to shut down.... done
server stopped
[postgres@node1 data]$ [2020-09-29 17:22:34] [WARNING] unable to ping "host=192.168.5.132 port=5432 user=postgres  dbname=postgres"
[2020-09-29 17:22:34] [DETAIL] PQping() returned "PQPING_NO_RESPONSE"
[2020-09-29 17:22:34] [WARNING] connection to node "node1" (ID: 1) lost
[2020-09-29 17:22:34] [DETAIL] 
FATAL:  terminating connection due to administrator command
server closed the connection unexpectedly
        This probably means the server terminated abnormally
        before or while processing the request.

[2020-09-29 17:22:34] [INFO] attempting to reconnect to node "node1" (ID: 1)
[2020-09-29 17:22:34] [ERROR] connection to database failed
[2020-09-29 17:22:34] [DETAIL] 
could not connect to server: Connection refused
        Is the server running on host "192.168.5.132" and accepting
        TCP/IP connections on port 5432?

[2020-09-29 17:22:34] [DETAIL] attempted to connect using:
  user=postgres dbname=postgres host=192.168.5.132 port=5432 connect_timeout=2 fallback_application_name=repmgr
[2020-09-29 17:22:34] [WARNING] reconnection to node "node1" (ID: 1) failed
[2020-09-29 17:22:34] [WARNING] unable to connect to local node
[2020-09-29 17:22:34] [INFO] checking state of node 1, 1 of 6 attempts
[2020-09-29 17:22:34] [WARNING] unable to ping "user=postgres dbname=postgres host=192.168.5.132 port=5432 connect_timeout=2 fallback_application_name=repmgr"

錯誤會一直報。


備庫查詢狀態，已經切換過來，時間可以設置重試、確認的時間和次數，默認6次，每次8s間隔：
[postgres@localhost pgsql-12]$ repmgr daemon status
 ID | Name  | Role    | Status    | Upstream | repmgrd | PID   | Paused? | Upstream last seen
----+-------+---------+-----------+----------+---------+-------+---------+--------------------
 1  | node1 | primary | - failed  | ?        | n/a     | n/a   | n/a     | n/a                
 2  | node2 | primary | * running |          | running | 79388 | no      | n/a 

啟動主庫，變為雙主：
[postgres@localhost data]$ repmgr daemon status
 ID | Name  | Role    | Status    | Upstream | repmgrd | PID   | Paused? | Upstream last seen
----+-------+---------+-----------+----------+---------+-------+---------+--------------------
 1  | node1 | primary | ! running |          | running | 16778 | no      | n/a                
 2  | node2 | primary | * running |          | running | 79388 | no      | n/a 
 

此時停止主節點，使用node rejoin的方式加入：

需要在repmgr.conf中添加配置，才能啟動數據庫，不需要重啟repmgrd：
service_start_command  = '/usr/pgsql-12/bin/pg_ctl -D /usr/pgsql-12/data start'
service_stop_command    = '/usr/pgsql-12/bin/pg_ctl -D /usr/pgsql-12/data stop'
service_restart_command = '/usr/pgsql-12/bin/pg_ctl -D /usr/pgsql-12/data restart'
service_reload_command  = '/usr/pgsql-12/bin/pg_ctl -D /usr/pgsql-12/data reload' 

--先進行rewind
repmgr node rejoin -h 192.168.5.132 -U postgres -d postgres --force-rewind --dry-run --verbose
--然后執行加入
repmgr node rejoin -h 192.168.5.132 -U postgres -d postgres --force-rewind --verbose

[postgres@node1 pgsql-12]$ repmgr node rejoin -h192.168.5.133 -Upostgres -dpostgres --force-rewind --verbose
INFO: checking for package configuration file "/etc/repmgr/12/repmgr.conf"
INFO: configuration file found at: "/etc/repmgr/12/repmgr.conf"
INFO: prerequisites for using pg_rewind are met
INFO: 0 files copied to "/tmp/repmgr-config-archive-node1"
NOTICE: executing pg_rewind
DETAIL: pg_rewind command is "pg_rewind -D '/usr/pgsql-12/data' --source-server='host=192.168.5.133 port=5432 user=postgres  dbname=postgres'"
pg_rewind: servers diverged at WAL location 0/3009918 on timeline 1
pg_rewind: no rewind required
NOTICE: 0 files copied to /usr/pgsql-12/data
INFO: directory "/tmp/repmgr-config-archive-node1" deleted
NOTICE: setting node 1's upstream to node 2
WARNING: unable to ping "host=192.168.5.132 port=5432 user=postgres  dbname=postgres"
DETAIL: PQping() returned "PQPING_NO_RESPONSE"
NOTICE: starting server using "/usr/pgsql-12/bin/pg_ctl -D /usr/pgsql-12/data start"
INFO: node "node1" (ID: 1) is pingable
INFO: node "node1" (ID: 1) has attached to its upstream node
NOTICE: NODE REJOIN successful
DETAIL: node 1 is now attached to node 2

[postgres@localhost pgsql-12]$ repmgr daemon status
 ID | Name  | Role    | Status    | Upstream | repmgrd | PID   | Paused? | Upstream last seen
----+-------+---------+-----------+----------+---------+-------+---------+--------------------
 1  | node1 | standby |   running | node2    | running | 27253 | no      | 0 second(s) ago    
 2  | node2 | primary | * running |          | running | 86308 | no      | n/a

5）當備機離線之后，再次加入，可能存在inactive狀態，怎么處理？

查看集群狀態：

[kingbase@node1 bin]$ repmgr cluster show
 ID | Name    | Role    | Status    | Upstream | Location | Priority | Timeline | Connection string                                                                                                                               
----+---------+---------+-----------+----------+----------+----------+----------+--------------------------------------------------------------------------------------------------------------------------------------------------
 1  | node132 | primary | * running |          | default  | 100      | 1        | host=192.168.5.132 user=esrep dbname=esrep port=54329 connect_timeout=3 keepalives=1 keepalives_idle=10 keepalives_interval=1 keepalives_count=3
 2  | node133 | standby | ! running | node132  | default  | 100      | 1        | host=192.168.5.133 user=esrep dbname=esrep port=54329 connect_timeout=3 keepalives=1 keepalives_idle=10 keepalives_interval=1 keepalives_count=3

WARNING: following issues were detected
  - node "node133" (ID: 2) is running but the repmgr node record is inactive

查看備機的rpmgr日志：

[2020-11-21 04:17:06] [NOTICE] repmgrd (repmgrd 5.0.0) starting up
[2020-11-21 04:17:06] [INFO] connecting to database "host=192.168.5.133 user=esrep dbname=esrep port=54329 connect_timeout=3 keepalives=1 keepalives_idle=10 keepalives_interval=1 keepalives_count=3"
[2020-11-21 04:17:06] [ERROR] this node is marked as inactive and cannot be used as a failover target
[2020-11-21 04:17:06] [HINT] Check that "repmgr (primary|standby) register" was executed for this node
[2020-11-21 04:17:06] [INFO] repmgrd terminating...

看日志可以發現需要重新注冊備機，在備機重新注冊：

[kingbase@localhost kingbase]$ repmgr standby register -F
INFO: connecting to local node "node133" (ID: 2)
INFO: connecting to primary database
INFO: standby registration complete
NOTICE: standby node "node133" (ID: 2) successfully registered

再從主節點看集群狀態，已經正常：

[kingbase@node1 bin]$ repmgr cluster show
 ID | Name    | Role    | Status    | Upstream | Location | Priority | Timeline | Connection string                                                                                                                               
----+---------+---------+-----------+----------+----------+----------+----------+--------------------------------------------------------------------------------------------------------------------------------------------------
 1  | node132 | primary | * running |          | default  | 100      | 1        | host=192.168.5.132 user=esrep dbname=esrep port=54329 connect_timeout=3 keepalives=1 keepalives_idle=10 keepalives_interval=1 keepalives_count=3
 2  | node133 | standby |   running | node132  | default  | 100      | 1        | host=192.168.5.133 user=esrep dbname=esrep port=54329 connect_timeout=3 keepalives=1 keepalives_idle=10 keepalives_interval=1 keepalives_count=3

6）如果沒有配置觀察節點，讓集群發生了腦裂，怎么處理？

比如網絡原因，主節點網絡隔離，備節點認為主節點離線了，會被提升為主庫（即使配置了turst_server，如果整個主節點網絡隔離了也會ping不通，生產中，如果網絡很差，可以把超時時間設置長一點）；網絡恢復后，集群會出現兩個primary。這種情況，首先要備份兩個主節點的數據，然后就想辦法先把集群先恢復，恢復時可以考慮以哪個主為新的主節點。

這里建議的方法：

A）以timeline較高的節點為新主節點，這樣另外一個主節點重新regioin進集群即可。操作方法：將時間線低的節點停止，在主節點執行剔除這個節點id，然后啟動這個節點，最后執行重新加入集群的命令即可：

新主節點執行：
repmgr primary unregister --node-id n

節點重啟后，重新加入集群：
repmgr node rejoin -h 192.168.5.132 -U postgres -d postgres

如果有日志分支，則需要加上 --force-rewind來消除。

B) 以timeline低的節點為新主節點，則需要重做timeline高的節點，參照新建備機的操作。

最后，A、B兩種方式恢復集群后，再根據備份過的data目錄，通過查詢數據、應用摘取等方式，將差異數據合並到新的集群中。

免責聲明！

本站轉載的文章為個人學習借鑒使用，本站對版權不負任何法律責任。如果侵犯了您的隱私權益，請聯系本站郵箱yoyou2525@163.com刪除。

猜您在找 repmgr的見證節點和守護進程 repmgr學習記錄（主從切換） repmgr+pg12集群，掛掉的主機如何手動加入集群 lightdb/postgresql高可用之repmgr日常管理及異常排查指南 Gradle基本使用（1）：安裝、IDEA使用 pycharm安裝和首次使用使用Homebrew安裝MySQL Graphviz安裝及簡單使用 PHPExcel安裝及使用 gmp安裝及使用摘要