Recover Failing pod

If you are running OneDB as an HA cluster. A primary and secondary and you have a failure of a single pod that doesn’t recover, you don’t need to perform a restore. Instead you can recover only the failing pod.

To force the pod to be recovered set an annotation to start the recovery:
kubectl annotate pod onedb-server-1 onedb_force_ifxclone=true –overwrite=true
Once the pod has successfully be cloned disable this annotation:
kubectl annotate pod onedb-server-1 onedb_force_ifxclone=false –overwrite=true
Note: This shows a recovery of pod onedb-server-0.