redis节点失联，日志：FAIL message received from

解决办法：

1.调整主节点client-output-buffer-limit参数，防止同步进行到一半失败

config set client-output-buffer-limit "slave 0 0 0"

2.调整cluster-node-timeout，不能少于15s

config set cluster-node-timeout 15000

FAIL message received from错误及md=psync scheduled to be closed ASAP for overcoming of output buffer limits错误

3.禁止任何耗时超过cluster-node-timeout的慢查询，因为会导致主从切换

参考：

redis集群故障分析

持久化导致内存阻塞或者主从通信超时