Allow the ability for a rolling cluster restart like during maintenance
Currently there is no way to test how applications behave during maintenance apart from waiting for maintenance, the ability to force a rolling cluster restart as happens during maintenance would be great. The failover is a different process and affects primary mainly.
17
votes
![](https://secure.gravatar.com/avatar/c539d80793627dd57cdac3d5d1dec0ae?size=40&default=https%3A%2F%2Fassets.uvcdn.com%2Fpkg%2Fadmin%2Ficons%2Fuser_70-6bcf9e08938533adb9bac95c3e487cb2a6d4a32f890ca6fdc82e3072e0ea0368.png)
-
Sergey commented
+1
We must run "service continuity test" for every application on scheduled basis. To do this we must shutdown nodes on demand to demonstrate that application can survive it.