Hyper-V backups can be stored locally and in the cloud. When Hyper-V backups are stored locally, they are stored in a proprietary format (encrypted and compressed) and require a restore to access backed-up data.
When the backup agent is installed on the Hyper-V host, it can back up any virtual machine on that host. You can select any number of virtual machines to be backed up. However, a VM can only be included in one backup set at a time.
If you are backing up to the cloud, then ensure that each VM you need to back up has the most recent version of Integration Services (IS) installed. The Hyper-V host’s Integration Services also must be updated (if the host’s IS is older, then the backup fails).
You can schedule a backup set to run on a recurring schedule. The default is running once per night. See Backup Schedules for more information.
You can add more than one VM to a backup set, but, by default, only one virtual disk is processed at a time when the backup runs. You can change this behavior by increasing the number of concurrent Hyper-V worker threads on the System Page. Increasing the number of worker threads allows multiple disks to be backed up concurrently but may use more resources on the host.
Hyper-V backups copy a full, initial backup of selected VMs. After this initial backup, incremental backups containing only the changed portions of each VM are performed until the next full backup is scheduled.