kubegres
kubegres copied to clipboard
Feature Request - Failover without new Replication
Would be nice if you could implement Failover without creating new Pods. So if primary fails it selects randomly one of the replication nodes and promotes it as primary until original primary node is back up.
That seems very logical to me - I don’t need another statefulset/pod when I already have a few to spare.
exactly
It seems I was needing the same thing here:
https://github.com/reactive-tech/kubegres/issues/88#issuecomment-1040351243