redis故障修复

1.redis集群扩容分配槽过程被打断。

[root@mongo-01 src]# ./redis-trib.rb reshard 10.0.0.114:6380

image.png

[root@mongo-01 src]# ./redis-trib.rb reshard 10.0.0.114:6380

image.png
image.png image.png

登录两个库查看发现异常故执行fix修复命令

[root@mongo-01 src]# ./redis-trib.rb fix 10.0.0.114:6380

image.png

[root@mongo-01 src]# ./redis-trib.rb info 10.0.0.114:6380
10.0.0.114:6380 (349cda52...) -> 0 keys | 5461 slots | 1 slaves.
10.0.0.116:6380 (bb3becf6...) -> 0 keys | 5461 slots | 1 slaves.
10.0.0.114:6391 (1c755430...) -> 0 keys | 0 slots | 0 slaves.
10.0.0.114:6390 (3cc289c6...) -> 0 keys | 994 slots | 0 slaves.
10.0.0.115:6380 (d69f195d...) -> 0 keys | 4468 slots | 1 slaves.
[OK] 0 keys in 5 masters.
0.00 keys per slot on average.

[root@mongo-01 src]# ./redis-trib.rb check 10.0.0.114:6380

image.png

oot@mongo-01 src]# ./redis-trib.rb rebalance 10.0.0.114:6380
[root@mongo-01 src]# ./redis-trib.rb info 10.0.0.114:6380
10.0.0.114:6380 (349cda52...) -> 0 keys | 4096 slots | 1 slaves.
10.0.0.116:6380 (bb3becf6...) -> 0 keys | 4096 slots | 1 slaves.
10.0.0.114:6391 (1c755430...) -> 0 keys | 0 slots | 0 slaves.
10.0.0.114:6390 (3cc289c6...) -> 0 keys | 4096 slots | 0 slaves.
10.0.0.115:6380 (d69f195d...) -> 0 keys | 4096 slots | 1 slaves.
[OK] 0 keys in 5 masters.
0.00 keys per slot on average.

修复成功！！！！

解决方法二

1.在两个问题节点执行
CLUSTER SETSLOT <slot> STABLE 命令
取消对槽 slot 的导入（import）或者迁移（migrate）。
2.这个时候发现槽同时存在两个库上，只能选择丢弃这个槽内的数据，在两个节点都卸载这个槽
CLUSTER DELSLOTS <slot> [slot ...]
3.CLUSTER ADDSLOTS <slot> [slot ...]，将卸载的槽添加至集群，集群恢复正常