Backup and restore of virtual machines

IBM Fusion protects virtual machines from data loss and corruption by backing up the virtual machines regularly. The virtual machines can then be restored from a previous backup.

Backup virtual machines

Backup policies define the frequency of backups, their storage locations, and the duration for which they must be retained. A best practice is to store backups in an object storage outside of the OpenShift® cluster, as this allows virtual machines to be recovered even when the original cluster is lost or compromised.

Virtual machine backups are incremental, meaning that only changed data is part of each backup. This approach reduces the amount of network traffic that is needed to transfer backups to object storage, and reduces the object storage requirements.

The IBM Fusion offers the following virtual machine backups through the Backup & Restore service.

In-place backup of virtual machine and Red Hat® OpenShift resources
Virtual machines can be restored in place, or in a different Red Hat OpenShift cluster.
Option to copy backup to object storage

Backups can be stored in object storage that is on a site or in the public cloud.

Integrated backup policy management
Virtual machines are protected by using backup policies, which define the frequency of backups, their storage locations, and the duration for which they must be retained.
Incremental backup of virtual machine
Only changes are included in each backup, reducing the amount of network traffic and object storage utilization.
To protect a group of virtual machines that represents an application, assign a backup policy to the namespace that the virtual machines belong to. It automatically backs up all the virtual machines in the namespace that uses the backup policy.
Note: A Red Hat OpenShift best practice is to create a namespace for a collection of virtual machines that represent an individual application.

To assign backup policies to an application that contains the virtual machines, go to the Applications page in the IBM Fusion user interface. Select the namespace that the virtual machines belong to and then assign a backup policy. Alternatively, you can manage backup policies declaratively by using the Custom Resources. It enables you to manage a large fleet of applications by using a GitOps approach, where a PolicyAssignment CR is deployed to the cluster alongside the application.

Restore virtual machines

To recover an application that is made up of virtual machines, select a backup to restore. Use the timestamp of each backup to select the appropriate point in time for restoration. The application can be restored in place, or you can restore the application to a different Red Hat OpenShift cluster.

IBM Fusion provides the following recovery options through its Backup & Restore service:
  • You can recover a virtual machine to the same or alternative cluster.
  • You can recover all virtual machines in a namespace to an alternative namespace and then delete the virtual machines that are no longer needed.
  • You can recover all virtual machines in a namespace to an alternative namespace and then copy files from the target virtual machine by using the command line.
  • You can manage the assignment of a backup policy to a high volume of virtual machines by directly working with the Fusion data protection custom resources, specifically the resources that are used to create backup policies and backup policy assignments.
  • You can restore subsets of virtual machines that belong to an application by using the following two methods:
    • You can restore the entire application into an alternative namespace and then delete the virtual machines that are no longer needed.
    • You can restore the entire application into a namespace to an alternative namespace and then copy files from the target virtual machine by using the command line.
To validate, restart the virtual machine after recovery.