Host-level Hyper-V backups | Why we back up each virtual machine separately
Why we back up each virtual machine separately
We get asked this a lot: "Why not just back up the Hyper-V host instead of every virtual machine?"
Fewer agents. Fewer licences. One job to manage. It sounds like a sensible shortcut - and if budget is tight, the cost difference is real.
But the backup is not the problem. Recovery is.
Yes, host-level backups cost less. Here is what that buys you.
Back up the Hyper-V host instead of each guest machine and you cut licence count. That saving lands in the budget and feels clean on paper.
What it does not buy you is a fast, clean, tested route back to running systems when something breaks. And something will break - ransomware, hardware failure, a botched update, an accidental deletion on the wrong day.
That is the day the backup gets judged. Not the day it was configured.
How host-level backups actually work
Hyper-V has a VSS writer. Backup software uses it to snapshot the host machine - virtual hard disk files included - while guest servers keep running.
One agent. Multiple VMs captured in a single pass. The files land in the backup. Looks complete.
It is not complete. It is a copy of disk images with no practical way to start them.
Four reasons we do not use host-level backups
Not edge cases. Not theoretical. These are the exact problems that surface when an organisation is down, staff cannot work, and someone needs answers fast.
1. You cannot spin up the VMs
Back up each VM individually and you can virtualise it directly - on a local appliance or in the cloud - within minutes of a failure.
Host-level backup gives you VHD and VHDX files. But backup appliances and cloud recovery environments do not support nested virtualisation. You cannot start those guest machines during a disaster. The data exists. The systems do not run. That distinction matters enormously at 2am on a Sunday.
2. Restoring a single file is a project
Someone deletes a folder. A database corrupts. Three days of work needs pulling back from Tuesday's backup.
With individual VM backups, that takes minutes. With a host-level backup, you download the entire virtual disk image, mount it somewhere, then dig out the file. A routine restore turns into a half-day task. That cost adds up fast.
3. AutoVerify cannot check the systems that matter
Every night, our backup platform boots each protected system and confirms it starts. That nightly check is how you know the backup will actually work - not just that it ran.
Protect only the Hyper-V host and that check validates one thing: the host. Not the file server. Not the line-of-business application. Not the domain controller. The systems your organisation depends on go unverified.
4. Disasters do not schedule themselves
Ransomware hits on a Thursday afternoon. Hardware fails on a bank holiday weekend. The call comes in when your team is already stretched.
Nobody discovers a backup limitation at a convenient moment. We would rather surface it now, fix it before it matters, and remove it from the list of things that can go wrong.
Who individual VM backups are right for
- Businesses running critical systems on Hyper-V
- Schools and multi-site organisations where downtime stops work
- Care homes, charities and professional services with continuity or compliance obligations
- Any organisation where one server going down has a measurable cost
- Anyone who needs to know the backup works - not just that it ran
Who host-level backups are not right for
- Any organisation that needs to recover fast from ransomware or hardware failure
- Anyone using cloud or appliance virtualisation as their recovery route
- Businesses that need quick, routine file and folder restores
- Organisations that want each server independently verified every night
- Any environment where downtime has a cost - financial, operational or reputational
We would not sell a host-level backup to a client whose systems genuinely matter to their business. The licence saving does not hold up against what you give away.
Our position: Anyone can take a backup. Recovery is the hard part - and the only part that counts. We protect each server individually so that when something fails, the route back is fast, tested and already proven to work.
What we do instead
We install the backup agent inside each virtual machine. Not on the host.
That gives every protected system its own recovery path:
- instant virtualisation on a local appliance or in the cloud
- file and folder restores that take minutes, not hours
- nightly boot verification for each individual server
- full flexibility when disaster recovery goes live under pressure
- documented proof that each system has been tested and will start
Yes, it means more licences. It also means that when something breaks - and eventually something will - every system on that list can come back.
The short version
Host-level Hyper-V backups are cheaper. They are also harder to recover from, impossible to properly verify, and useless for instant virtualisation.
Cheaper backup is not the same as better protection. For most businesses, schools and organisations, the difference only becomes obvious at the worst possible moment.
We would rather you never find out the hard way.
Looking for a backup solution?
Whether you need to protect on-premise servers, Hyper-V virtual machines, endpoints or Microsoft 365 data, we can help you choose a backup strategy that focuses on recovery, not just retention.
Axcient Server & Endpoint Backup
Protect physical servers, Hyper-V virtual machines, workstations and critical business data with rapid recovery options, cloud failover and automated verification.
Microsoft 365 Cloud Backup
Protect Exchange Online, OneDrive, SharePoint and Microsoft Teams data from accidental deletion, ransomware and retention gaps.
Not sure if your backups would actually recover?
We will review your current setup, explain what we find in plain English, and tell you honestly whether your systems are protected - or just backed up.
Book a backup review