Openebs: Automatically move replica from the failing node

[ytsarev]

Description

Currently in case of long term node failure jiva replicas have to be moved manually. It would be great if some automation(operator?) would detect node failure and move replica to the healthy node.

Context

From time to time we experience node failures where jiva replicas do reside.
In that case we have to options - either recover the node or move replica to the healthy one.
In second case we have to manually edit node affinity of replica deployment so that it is scheduled on healthy node.

Possible Solution

Operator/controller that detects node failure and moves replica to healthy node automatically.

This issue is reported for jiva due to recent operation experience but i believe the similar style move of cstor-pool pods will be also extremely useful

[utkarshmani1997]

Thanks @ytsarev for filing this feature request, we are targeting this feature in upcoming releases.

[sbaryakov]

+1

[kmova]

Taking this up for 1.4 release. An alpha version is being targeted for 1.3

[github-actions[bot]]

Issues go stale after 30d of inactivity.

[github-actions[bot]]

Issues go stale after 90d of inactivity.