本文共 3037 字，大约阅读时间需要 10 分钟。
实验环境
| 名称 | IP | 系统 |
| Master | 172.17.10.190 | Centos 6.5 |
| Slave | 172.17.10.189 | Centos 6.5 |
1.yun安装
| 1 2 | rpm -ivh https: //download .postgresql.org /pub/repos/yum/9 .6 /redhat/rhel-6-x86_64/pgdg-centos96-9 .6-3.noarch.rpm yum install postgresql96.x86_64 postgresql96-server.x86_64 -y |
2.主从配置
2.1 主数据库配置
启动master
| 1 2 3 4 | /etc/init .d /postgresql-9 .6 initdb /etc/init .d /postgresql-9 .6 start su - postgres psql |
授权
| 1 | create role repl login replication encrypted password '51idc.com' ; |
编辑hba文件
/var/lib/pgsql/9.6/data/pg_hba.conf
新增
| 1 2 | host replication repl 172.17.10.0 /24 md5 host all repl 172.17.10.0 /24 trust |
编辑配置文件
/var/lib/pgsql/9.6/data/postgresql.conf
| 1 2 3 4 5 6 7 8 | listen_addresses = 172.17.10.190 wal_level = hot_standby #热备模式 max_wal_senders= 6 #可以设置最多几个流复制链接，差不多有几个从，就设置多少 wal_keep_segments = 10240 #重要配置 wal_send_timeout = 60s max_connections = 512 #从库的 max_connections要大于主库 archive_mode = on #允许归档 archive_command = 'cp %p /url/path%f' #根据实际情况设置 |
2.2 从数据库配置
| 1 | su - postgres |
如果开始为启动数据库可忽略下一步
| 1 2 3 | rm -rf /var/lib/pgsql/9 .6 /data/ * #开始没有启动从数据库，这一步可以省略 pg_basebackup -h 172.17.10.190 -U repl -D /var/lib/pgsql/9 .6 /data -X stream -P cp /usr/pgsql-9 .6 /share/recovery .conf.sample /var/lib/pgsql/9 .6 /data/recovery .conf |
修改配置文件recovery.conf
| 1 2 3 4 | standby_mode = on primary_conninfo = 'host=172.17.10.190 port=5432 user=repl password=51idc.com' trigger_file = '/var/lib/pgsql/9.6/data/trigger.kenyon' #主从切换时后的触发文件 recovery_target_timeline = 'latest' |
配置postgresql.conf文件
| 1 2 3 4 5 6 7 | listen_addresses = 172.17.10.189 wal_level = hot_standby max_connections = 1000 #一般从的最大链接要大于主的。 hot_standby = on #说明这台机器不仅仅用于数据归档，也用于查询 max_standby_streaming_delay = 30s wal_receiver_status_interval = 10s #多久向主报告一次从的状态。 hot_standby_feedback = on #如果有错误的数据复制，是否向主进行范例 |
检测
| 1 | select client_addr,sync_state from pg_stat_replication; |
查看主从状态
| 1 | select * from pg_stat_replication; |
脚本监控主从
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | #!/bin/bash # mail xuel@51idc.com data=` date +%Y-%M-%d " " %H:%m` netstat -lntup| grep 5432 && ps -ef| grep postmaster if [ $? - eq 0 ]; then for IP in 172.17.10.188 172.17.10.189 do /usr/bin/psql -h 172.17.10.190 -p 5432 -U repl -d postgres -- command "select * from pg_stat_replication" | grep $IP if [ "$?" != "0" ]; then echo "postgresql master-slave status is error! please login check!" |mail -r "xuel@51idc.com" -s "postgresql master-slave status is error" xuel@51idc.com \ && echo "$data postgresql postgresql master-slave status is error!" >> /var/log/postgresql-error .log fi done else echo "postgresql master-slave status is error! please login check!" |mail -r "xuel@51idc.com" -s "postgresql master-slave status is error" xuel@51idc.com \ && echo "$data postgresql postgresql master-slave status is error!" >> /var/log/postgresql-error .log fi |
2.3主从切换
主库查看进程为sender
备库
停止主库
查看slave的日志
创建触发文件，切换主
| 1 | touch trigger.kenyon |
查看slave的日志，面前已经切换为主
使用pg_controldata
备机状态为： in archive recovery
主库状态为：in production