Skip to content
AlpCloud
All posts

How to back up Hyper-V and VMware virtual machines

Agent inside the VM or hypervisor-level backup? Recovery options, retention planning and common mistakes in virtual machine backup.

Today many companies run their servers as virtual machines on Hyper-V or VMware vSphere rather than on physical hardware. Virtualisation makes operations easier, but if backup isn't designed properly, a single host failure can take down dozens of servers at once.

Two approaches: agent inside the VM or hypervisor level?

An agent inside each VM: Backup software is installed in every virtual machine, and each one is backed up like a physical server. This works for small environments but becomes hard to manage as the number of machines grows.

Hypervisor-level backup: Backups are taken through the host or vCenter that runs the virtual machines. There's no need to install an agent in every VM, and new machines are easy to add to the plan.

AlpCloud backs up VMware ESXi/vCenter and Microsoft Hyper-V environments at the hypervisor level.

Recovery options

A backup is only as good as how flexibly you can recover from it. With AlpCloud you can:

  • Restore an entire virtual machine to the same or a different host.
  • Use granular file restore to recover individual files and folders without restoring the whole VM.
  • Convert image backups of physical servers to VHD/VHDX or VMDK disks and run them as virtual machines — also a practical way to move an old physical server to a VM.
  • Bring image backups up as virtual machines in Amazon EC2 or Azure.

How to plan retention

Virtual machine backups are large; taking a full backup every day is expensive in both time and storage. Our recommended approach:

  1. Take a full backup the first time.
  2. On the following days, take incremental backups that send only changed blocks.
  3. For long-term retention, keep weekly and monthly copies for separate periods with the GFS model.
  4. Use the same plan to back up to a local NAS for fast recovery and to the cloud for disaster scenarios (hybrid backup).

Common mistakes

  • Protecting only the host: RAID or redundant disks are not a backup; they don't bring back deleted or encrypted data.
  • Keeping backups on the same host: When the host fails, the backup goes with it.
  • Treating snapshots as backups: Snapshots are short-term rollback points; they live on the same storage and hurt performance when kept for long.
  • Never testing a restore: Regularly restore a VM to a test environment to confirm your backups work.

Find out more on our VMware and Hyper-V backup page.

Secure your backups today

We'll set up your trial account and walk you through the setup.