redis-cluster-operator icon indicating copy to clipboard operation
redis-cluster-operator copied to clipboard

Published 20 hours ago • verified publisher icon ucloud

Redis cluster can not recover from failure

Open laoluotuo opened this issue 5 years ago • 6 comments

I have deployed a 3-master-3-slave Redis cluster in my stage environment and it is running well. But after I have deleted one of these six pods, the cluster can't recover. The IP address of the pod deleted before is still there, in "fail" status. So there are 7 nodes in my sight: 1 master failed, 1 master running but with no slots, 3 masters with slots, 2 slaves. How to get it back to normal state? Could you add a new feature to solve this problem? Thank you very much!

laoluotuo avatar Dec 27 '19 10:12 laoluotuo

@laoluotuo Now only the redis cluster that use persistent storage can recover after accidental deletion,I will fix this later.

polefishu avatar Jan 16 '20 06:01 polefishu

Doest that mean this operator is now production ready yet?

zedtux avatar Jan 22 '20 10:01 zedtux

Hi @polefishu can a redis cluster with out persistent storage recover from failure now ?

amrmahdi avatar Apr 03 '20 07:04 amrmahdi

Doest that mean this operator is now production ready yet?

I will say it is BETA version. It has been deployed in our production environment for weeks, basic features like are working well, but more tests are needed for "password change" and "backup/restore". We will release v1.0 when it is production ready.

gaopenghigh avatar Apr 08 '20 02:04 gaopenghigh

Is it ok now to use a cluster without persistent storage?

vmrm avatar Apr 30 '21 09:04 vmrm

Is there any updates or plans to add healing of clusters without persistent storage?

dm3ch avatar Feb 15 '22 21:02 dm3ch