NetBackup™ Deployment Guide for Amazon Elastic Kubernetes Services (EKS) Cluster

Perform the following steps to rollback from upgrade failure and install the NetBackup version prior to upgrade

1. Delete the environment CR object using the following command and wait until all the underlying resources are cleaned up:

   kubectl delete environment.netbackup.veritas.com <environment name> -n <namespace>

   For example, primary server CR, media server CR, MSDP CR and their underlined resources.

2. Delete the new operator which is deployed during upgrade using the following command:

   kubectl delete -k <new-operator-directory>

   This will delete the new operator and new CRDs.

3. Apply the NetBackup operator directory which was preserved (the directory which was used to install operator before upgrade) using the following command:

   kubectl apply -k <operator_directory>

4. Get names of PV attached to primary server PVC (data, catalog and log) using the following command:

   kubectl get pvc -n <namespace> -o wide

5. Delete the primary server PVC (data, catalog and log) using the following command:

   kubectl delete pvc <pvc-name> -n <namespace>

6. Delete the PV linked to primary server PVC using the following command:

   kubectl delete pv <pv-name> command

7. Edit the preserved environment.yaml file (from older version of NetBackup package directory) and remove keySecret section from MSDP Scaleout section. Also change the CR spec paused: false to paused: true for every object in MSDP Scaleout and media servers section.
8. Apply the edited environment.yaml file using the following command:

   kubectl apply -f <environment.yaml>

9. After the primary server pod is in ready state (1/1), change the CR spec from paused: false to paused: true in environment.yaml file of the primary server section and reapply the environment.yaml using the following command:

   kubectl apply -f environment.yaml -n <namespace>

10. Exec into the primary server pod using the following command:

    kubectl exec -it -n <PrimaryServer/MediaServer-CR-namespace> <primary-pod-name> -- /bin/bash

    • Increase the debug logs level on primary server.

    • Create a DRPackages directory at the persisted location using mkdir /mnt/nblogs/DRPackages folder.

    • Change ownership of the DRPackages folder to service user using the following command:

      chown nbsvcusr:nbsvcusr /mnt/nblogs/DRPackages

11. Copy the earlier copied DR files to primary pod at /mnt/nblogs/DRPackages using the following command:

    kubectl cp <Path_of_DRPackages_on_host_machine> <primary-pod-namespace>/<primary-pod-name>:/mnt/nblogs/DRPackages

12. Execute the following steps in the primary server pod:

    • Change ownership of the files in /mnt/nblogs/DRPackages using the following command:

      chown nbsvcusr:nbsvcusr <filename>

    • Deactivate NetBackup health probes using the following command:

      /opt/veritas/vxapp-manage/nbu-health deactivate

    • Stop the NetBackup services using the following command:

      /usr/openv/netbackup/bin/bp.kill_all

    • Execute the following command:

      nbhostidentity -import -infile /mnt/nblogs/DRPackages/<filename>.drpkg

    • Restart all the NetBackup services using the following command:

      /usr/openv/netbackup/bin/bp.start

13. Verify if the security settings are enabled.
14. Add respective media server entry in host properties using NetBackupAdministration Console as follows:

    Navigate to NetBackup Management > Host properties > Master Server > Add Additional server and add media server.

15. Restart the NetBackup services in primary server pod and external media server as follows:

    • Exec into the primary server pod using command:

      kubectl exec -it -n <PrimaryServer/MediaServer-CR-namespace> <primary-pod-name> -- /bin/bash

    • Run the following command to stop all the services:

      /usr/openv/netbackup/bin/bp.kill_all

      After stopping all the services, restart the services using the following command:

      /usr/openv/netbackup/bin/bp.start_all

    • Run the following command to stop all the NetBackup services:

      /usr/openv/netbackup/bin/bp.kill_all

      After stopping all the services, restart the NetBackup services using the following command:

      /usr/openv/netbackup/bin/bp.start_all

16. Configure a storage unit on external media server that is used during catalog backup.
17. Perform catalog recovery from NetBackup Administration Console.

    For more information, refer to the Veritas^TM NetBackup Troubleshooting Guide

18. Exec into the primary server pod using the following command:

    kubectl exec -it -n <PrimaryServer/MediaServer-CR-namespace> <primary-pod-name> -- /bin/bash

    • Stop the NetBackup services using the following command:

      /usr/openv/netbackup/bin/bp.kill_all

    • Start the NetBackup services using the following command:

      /usr/openv/netbackup/bin/bp.start_all

    • Activate NetBackup health probes using the following command:

      /opt/veritas/vxapp-manage/nbu-health activate

19. Restart the NetBackup operator pod, where user must delete the pod using the following command:

    kuebctl delete <operator-pod-name> -n <namespace>

    Kubernetes will start new pod after deletion.

20. Pause the reconciler for primary, media servers, and msdp scaleouts in the following sequence:

    • Change CR spec paused: true to paused: false in environment.yaml file of the primary section and re-apply environment.yaml file using the following command:

      kubectl apply -f environment.yaml -n <namespace>

      Wait till primary server is in ready state.

    • Change CR spec paused: true to paused: false in environment.yaml file of the msdp scaleouts section and re-apply environment.yaml file using the following command:

      kubectl apply -f environment.yaml -n <namespace>

      Wait till primary server is in ready state.

    • Change CR spec paused: true to paused: false in environment.yaml file of the media servers section and re-apply environment.yaml file using the following command:

      kubectl apply -f environment.yaml -n <namespace>

      Wait till primary server is in ready state.

21. Verify the rollback is successful by performing backups and recovery jobs.