(1)9 shard,3 node
(2)master node宕机,自动master选举,red
(3)replica容错:新master将replica提升为primary shard,yellow
(4)重启宕机node,master copy replica到该node,使用原有的shard并同步宕机后的修改,green
(1)9 shard,3 node
(2)master node宕机,自动master选举,red
(3)replica容错:新master将replica提升为primary shard,yellow
(4)重启宕机node,master copy replica到该node,使用原有的shard并同步宕机后的修改,green