Move managed clusters using the ACM 2.12 backup component

Learn how to use the hub cluster backup and restore operator to move managed clusters from one hub to another using Red Hat Advanced Cluster Management for Kubernetes 2.12. 

Try Red Hat Advanced Cluster Management for KubernetesLearn more

With the primary hub ready to go, let’s now prepare the secondary hub.

Prerequisites:

  • Understand how the backup and restore operation works. 
  • Knowledge of how to install a hub cluster.
  • The ability to enable the backup and restore component.

In this lesson, you will:

  • Pause the backup schedule.
  • Update the DataProtectionApplication storage location.
  • Ensure the backups sync. 

Pause the backup schedule

If Hub 2 has a BackupSchedule resource in Enabled state, pause it by setting the paused property to true on the BackupSchedule (Figure 1). The BackupSchedule resource has the paused property set to true

The backup schedule in Enabled state with Velero schedules enabled.
Figure 1: The backup schedule in Enabled state with Velero schedules enabled.

You need to pause the backup schedule because in the next section, you must change the storage location to point to the Hub 1 storage location. That way, you can access the Hub 1 backup data. 

If you don't pause the BackupSchedule on Hub 2, it will create Hub 2 backups at the Hub 1 storage location. You don’t want to back up data from Hub 2 during this process, since you are using the Hub 1 storage location.

Update the DataProtectionApplication storage location

Update the DataProtectionApplication resource on Hub 2 to point to the Hub 1 storage location. By doing this, Hub 2 can access the backups created by Hub 1. 

For example, see the following YAML sample where the DataProtectionApplication resource on Hub 2 points to the Hub 1 storage location: 

apiVersion: oadp.openshift.io/v1alpha1
kind: DataProtectionApplication
metadata:
 name: dpa-hub-2
 namespace: open-cluster-management-backup
spec:
 backupLocations:
   - velero:
       default: true
       objectStorage:
         bucket: vb-velero-backup
         prefix: hub-1 # this is Hub1 storage location
```

Ensure the backups sync 

Ensure that the Velero BackupStorageLocation on Hub 2 has the label Phase status set to Available. Verify that you can see the latest backups created on Hub 1.  Figure 2 shows an example of this. 

BackupStorageLocation is available.
Figure 2: BackupStorageLocation is available.

Verify that the backups you see on Hub 2 are created by Hub 1 by looking at the clusterId labels for the backup resources. The value of the clusterId label annotation should be the clusterId for Hub 1, as shown in Figure 3.

Hub 1 backups synced and available on Hub 2.
Figure 3: Hub 1 backups synced and available on Hub 2.

Now that we’ve set up both hubs, we can move on to the final phase of this process: moving managed clusters between the two hubs.

Previous resource
Prepare the primary hub
Next resource
Run the restore operation on Hub 2