VM Backup

To produce backups of VMs, Veeam Plug-in for Scale Computing HyperCore runs backup jobs. A backup job is a collection of settings that define the way backup operations are performed: what data to back up, where to store backups, when to start the backup process, and so on.

Veeam Plug-in for Scale Computing HyperCore does not install agent software inside VMs to back up VM data — it uses native Scale Computing HyperCore capabilities instead. During every backup session, Veeam Plug-in for Scale Computing HyperCore creates a Scale Computing HyperCore copy-on-write snapshot of each VM added to a backup job. The snapshot is further used to create a VM backup.

How to Protect VMs

How VM Backup Works

Veeam Plug-in for Scale Computing HyperCore performs VM backup in the following way:

Launches a worker on the same cluster where the processed VM resides.

If no worker is deployed on the host, Veeam Plug-in for Scale Computing HyperCore launches a worker that is deployed on any other Scale Computing HyperCore host connected to the backup infrastructure.

Connects to the Scale Computing HyperCore cluster and creates a snapshot of the processed VM.
Uses the worker to read data from disks that are attached to the processed VM, compares it to the data written to the snapshot created at the step 2, excludes the changes and transfers the resulting data to the target repository — and stores it in the native Veeam format.

To reduce the amount of data read from VM disks, Veeam Plug-in for Scale Computing HyperCore uses the changed block tracking (CBT) mechanism: during incremental backup sessions, Veeam Plug-in for Scale Computing HyperCore compares the current disk content with the backed-up content and reads only those data blocks that have changed since the previous backup session. If CBT cannot be used, Veeam Plug-in for Scale Computing HyperCore reads all data from the VM disks. For more information, see Changed Block Tracking.

Veeam Plug-in for Scale Computing HyperCore compresses and deduplicates data saved to repositories.

Removes the previous snapshot and suspends the worker when the backup session completes.