Chinaunix

标题: keepalived+redis 主从，从挂掉恢复不能正常加入，必须要重启keepalived [打印本页]

作者: 梦落花香Z 时间: 2014-03-13 10:27
标题: keepalived+redis 主从，从挂掉恢复不能正常加入，必须要重启keepalived
先说一下我的环境
两台redis服务器分别为 192.168.5.234
                           192.168.5.236
用keepalived 实现主从切换。
                           192.168.5.234 和192.168.5.236依次启动，
                           5.234为主，5.236为辅
                           当5.234挂掉之后，5.236为主，接管业务
                           5.234启动为从
                           当5.236挂掉，5.234为主。

贴出一个keepalived的配置
vrrp_script chk_redis {
   script "/etc/keepalived/keepalived_scripts/redis_check.sh 127.0.0.1 6379"
   interval 1
}
vrrp_instance redis {
state BACKUP
interface eth0
virtual_router_id 51
priority 150
nopreempt
advert_int 1
authentication {
      auth_type PASS
      auth_pass redis6379
}
virtual_ipaddress {
      192.168.5.30
}
track_script {
      chk_redis
}
notify_master "/etc/keepalived/keepalived_scripts/redis_master_slave.sh 127.0.0.1 192.168.5.236 6379"
notify_backup "/etc/keepalived/keepalived_scripts/redis_backup_slave.sh 127.0.0.1 192.168.5.236 6379"
notify_fault  "/etc/keepalived/keepalived_scripts/redis_fault.sh"
notify_stop  "/etc/keepalived/keepalived_scripts/redis_stop.sh"
}

经测试，主从切换没有问题。
但是当我的从的redis挂掉了，修复以后启动并不能成为从，而是需要重启keepalived才能成为从。
我看了下keepalived的日志，有检测到从挂掉和启动了。但是没有执行任何notify 里面的脚本。notify里的脚本是有日志输出的。
请大家帮我分析下这里面的原因，谢谢了。

作者: q1208c 时间: 2014-03-13 11:14

　为什么越来越多的人用　keepalived　来切换7层的应用呢 .

作者: ding_cw 时间: 2014-03-28 14:44
state BACKUP 说明是从

virtual_router_id 51 是优先级别

一般需要设置主的优先级大于从

作者: 宁静致远_guan 时间: 2015-12-28 17:01
我想说，我的主从切换还有问题呢，不知道为什么从一直没法接管主的，一直报错[root@localhost scripts]# tailf /var/log/keepalived-redis-state.log
Run SLAVEOF cmd ...
ERR wrong number of arguments for 'slaveof' command

[master]
Mon Dec 28 16:42:46 CST 2015
Being master....
Run SLAVEOF cmd ...
Could not connect to Redis at -p:6379: Name or service not known
Run SLAVEOF NO ONE cmd ...
Could not connect to Redis at -p:6379: Name or service not known

欢迎光临 Chinaunix (http://bbs.chinaunix.net/)