Backup and Restore jobs fail with CRD errors

Summary

Some times backup and restore jobs fail with the following error in CloudCasa agent logs:

status.phase: Unsupported value: "Finalizing": supported values: "New", "FailedValidation", "InProgress",
  "WaitingForPluginOperations", "WaitingForPluginOperationsPartiallyFailed", "Completed", "PartiallyFailed",
  "Failed"

Solution

This error comes when the cluster has Velero CRDs that are not compatible with CloudCasa. To resolve the issue, follow these steps:

  • Confirm that there are CRDs on the cluster installed by Velero:

    kubectl get crd -lcomponent=velero

  • If the above command shows Velero CRDs, you need to delete them but it is very important to make sure that Velero is not being used on this cluster. If that is not the case and Velero is being used, please contact CloudCasa support for further steps. If Velero is not being used and you are OK to delete the CRDs, continue with the steps here.

  • Delete Velero CRDs:

    kubectl delete crd -lcomponent=velero

  • Delete cloudcasa-io namespace:

    kubectl delete ns cloudcasa-io

  • Find the CloudCasa agent URL by going to cluster’s Edit page in CloudCasa UI and clicking on “Install”.

  • Apply CloudCasa agent.