The slave node does not need to failover

[bbqccx]

When we find that the slave node is abnormally connected or the master node fails and switches to the slave, the slave node will still be added to the failover candidates and failover task process. In fact, the slave does not need to switch; it just needs to prompt for logs indicating the connection exception.

for example：
{"level":"info","timestamp":"2024-04-12T15:12:31.208+0800","caller":"probe/cluster.go:103","msg":"Add the node into the fail over candidates","id":"73o7UbE51haonhk5cr2IkRUVNPUlBIOlGDqam2Kb"
,"role":"slave","addr":"10.1.1.1:9012"}

{"level":"info","timestamp":"2024-04-12T15:12:25.215+0800","caller":"failover/cluster.go:215","msg":"Finish the fail over task","task":{"namespace":"nimtest","cluster":"nimtestkv-2","shard_
idx":6,"node":{"id":"73o7UbE51haonhk5cr2IkRUVNPUlBIOlGDqam2Kb","addr":"10.1.1.1:9012","role":"slave","password":"xxx","created_at":0},"type":1,"probe_count":0,"pending_time":1712905921,"start_time":1712905945,"finish_time":1712905942,"status":3,"error":""}}

[git-hulk]

@bbqccx Thanks for your feedback.