Restore from snapshot
NetBackup features a restore from snapshot function where you can restore all the backup images in a recovery point, using a single restore job. You can view the restore from snapshot job in the Activity monitor.
To restore from a snapshot
- On the left, click Workloads > Kubernetes.
- On the Namespace tab, click the namespace of the asset that you want to recover.
- Click the Recovery points tab.
The Recovery points tab shows you all the recovery points with the date, time, and copies of the backup. You can set filters to filter the displayed recovery points. Click the date in the Date column, to view the details of the recovery point. The Recovery points details dialog shows the resources that were backed up, like config maps, secrets, persistent volumes, pod, and so on. For details about these resources, see https://kubernetes.io/docs/reference/kubernetes-api/
- Locate the recovery point that you want to restore.
- In the Copies column, click the # copies button. For example, if there are two copies, the button displays as 2 copies.
- In the list of copies, locate the Snapshot copy. Then click Actions > Restore namespace.
Note:
Enable the Restore option for all infected copies by selecting Allow the selection of recovery points that are malware-affected option.
- In the Recovery target page, target cluster is auto populated.
Note:
Alternate cluster restore is not supported for snapshot copy.
- Under Specify destination namespace, select any of the following options for restore:
Use original namespace to use the original namespace that was backed up for restore. By default, this option is selected.
Use alternate namespace to use an alternate namespace for restore and then, click Next.
- Under Select resource types to recover, select any of the following resource types to restore:
All resource types to recover all resource types. By default, this option is selected.
Recover selected resource types to recover only the selected resource types.
Note:
Select resource types to recover option is for advance users. If you are not careful in selecting the resources that you want to restore, you may not get a fully functional namespace after restoring.
- Under Select Persistent volume claims to recover, select any of the following persistent volume claims to recover:
All Persistent volume claims to recover all persistent volume claims. By default, this option is selected.
Recover selected Persistent volume claims to recover selected persistent volume claims.
Note:
If you do not select any option in Recover selected resource types, then include empty persistent volume claims option is selected and no persistent volume claims is restored.
If you do not select any options in the Recover selected persistent volume claims then in the Recovery options section, it includes empty persistent volume claims and no persistent volume claims is restored.
Note:
Restore only persistent volume enables toggle in the selected persistent volume claims to restore only the persistent volume. This setting does not create a corresponding persistent volume claim.
- Click on the Failure strategy section to view the failure strategy options to recover.
- Under Select failure strategy to recover, select any of the following failure strategies to recover:
Note:
If a failure occurs during the restore of metadata or PVCs, the restore job runs according to the failure strategy selected.
Fail Fast to terminate the restore for any failure occurrence.
Proceed Ahead to continue restoring the next PVC. If the parent image (first image) restore fails then the restore job terminates.
Retry to specify a retry count for the metadata or PVC restore. If the restore fails even after retries, then the restore job terminates.
Note:
The selected failure strategy is displayed in the Activity monitor.
- Click Next.
- On the Recovery options page, click Start recovery to submit the recovery entry.
- In Activity monitor, click the Job ID to view the restore job details.
Note:
NetBackup Kubernetes restore uses single job to restore all the persistent volume claims and a namespace. You can view logs on the to track the restore of persistent volumes, persistent volume claims, or metadata.