Some basic support for group replication was previously added in #1180
However, it is only about understanding single primary groups and being able to display them in the UI, but it does not tackle
anything related to failure detection and recovery.

While group replication has built-in primary election and fail-over, this is only for the group itself. If any group member (be it the primary or any secondary) has traditional async/semi-sync replicas, and it fails, neither group replication nor orchestrator currently fix that.

Hence, this Issue is a feature request so that Orchestrator can fix traditional replicas of group members.