记一次Redis哨兵机制【主机挂了无法切换其他从机】的问题

• 日志打印

9873:X 05 Oct 18:32:56.967 # WARNING: The TCP backlog setting of 511 cannot be enforced because /proc/sys/net/core/somaxconn is set to the lower value of 128.
9873:X 05 Oct 18:32:56.970 # Sentinel ID is c3a7e3ace3b81031564a3dd1a590a234aa67c813
9873:X 05 Oct 18:32:56.970 # +monitor master host6379 127.0.0.1 6379 quorum 1
9873:X 05 Oct 18:33:26.989 # +sdown master host6379 127.0.0.1 6379
9873:X 05 Oct 18:33:26.989 # +odown master host6379 127.0.0.1 6379 #quorum 1/1
9873:X 05 Oct 18:33:26.989 # +new-epoch 1
9873:X 05 Oct 18:33:26.989 # +try-failover master host6379 127.0.0.1 6379
9873:X 05 Oct 18:33:26.993 # +vote-for-leader c3a7e3ace3b81031564a3dd1a590a234aa67c813 1
9873:X 05 Oct 18:33:26.993 # +elected-leader master host6379 127.0.0.1 6379
9873:X 05 Oct 18:33:26.993 # +failover-state-select-slave master host6379 127.0.0.1 6379
9873:X 05 Oct 18:33:27.045 # -failover-abort-no-good-slave master host6379 127.0.0.1 6379
9873:X 05 Oct 18:33:27.108 # Next failover delay: I will not start a failover before Sat Oct  5 18:39:27 2019
9873:X 05 Oct 18:39:27.833 # +new-epoch 2
9873:X 05 Oct 18:39:27.835 # +try-failover master host6379 127.0.0.1 6379
9873:X 05 Oct 18:39:27.838 # +vote-for-leader c3a7e3ace3b81031564a3dd1a590a234aa67c813 2
9873:X 05 Oct 18:39:27.838 # +elected-leader master host6379 127.0.0.1 6379
9873:X 05 Oct 18:39:27.838 # +failover-state-select-slave master host6379 127.0.0.1 6379
9873:X 05 Oct 18:39:27.905 # -failover-abort-no-good-slave master host6379 127.0.0.1 6379
9873:X 05 Oct 18:39:27.971 # Next failover delay: I will not start a failover before Sat Oct  5 18:45:28 2019

      这个问题我们可以分两种情况分析：

1. 主机的bind地址为0.0.0.0，而其他的主机极有可能bind地址为127.0.0.1，所以我们需要把所有从机的bind地址改为0.0.0.0即可。

• 从机redis配置文件（部分）图

• 将bind地址修改为0.0.0.0
  
  2.第二种情况是，因为从机也有密码，所以我们需要给定密码才能正确切换从机。

• 在主机的redis.conf文件中设置其他从机的密码
  

• 在哨兵模式核心配置文件中加入密码，主机与从机的密码需保持一致。
  * 经过测试以上方案都需要去尝试，终是解决了这个问题。