redis主从异常

1、现象

redis启用redis-sentinel架构，三台sentinel两台redis。发现从库没有从主库同步数据。

2、日志

31201:S 24 Feb 2023 14:26:59.626 * Connecting to MASTER 192.168.162.152:16380
31201:S 24 Feb 2023 14:26:59.626 * MASTER <-> REPLICA sync started
31201:S 24 Feb 2023 14:26:59.627 * Non blocking connect for SYNC fired the event.
31201:S 24 Feb 2023 14:26:59.627 * Master replied to PING, replication can continue...
31201:S 24 Feb 2023 14:26:59.627 # Unable to AUTH to MASTER: -ERR invalid password

3、查看slveof状态

从库，显示没连上master

127.0.0.1:16380> info replication
# Replication
role:slave
master_host:192.168.162.152
master_port:16380
master_link_status:down
master_last_io_seconds_ago:0
master_sync_in_progress:0
slave_repl_offset:1916743737
slave_priority:100
slave_read_only:1
connected_slaves:0
master_replid:9525edbaf58286234e4d0c2144a22a2cddd2d569
master_replid2:0000000000000000000000000000000000000000
master_repl_offset:1916743737
second_repl_offset:-1
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:1915695162
repl_backlog_histlen:1048576

主库：

用info replication命令查看，也没有salve连接

 

查看主从库的masterauth:

127.0.0.1:16380> CONFIG GET masterauth
1) "masterauth"
2) "ada68d"

主从库密码都是这个。

命令行手动去同步master也返回OK

redis 127.0.0.1:16380> SLAVEOF NO ONE
OK
redis 127.0.0.1:16380> SLAVEOF 192.168.162.152 16380
OK

 

4、最后用原来的配置新建了两个16379端口的redis主从去测试，发现把masterauth和requirepass密码设置成一样时，主从连接正常了。

https://blog.csdn.net/damanchen/article/details/100584275

说是：

主从在做复制的时候，如果有masterauth，从库会先发送masterauth给主库进行验证，验证无误后再进行下面的主从复制，验证不通过时就跳转到error错误处理。