What to Do When a VM Backup Freezes at the Same Percentage Only in Snapshot Mode

Eva Wong

IceWhale author

Eva Wong is the Technical Writer and resident tinkerer at ZimaSpace. A lifelong geek with a passion for homelabs and open-source software, she specializes in translating complex technical concepts into accessible, hands-on guides. Eva believes that self-hosting should be fun, not intimidating. Through her tutorials, she empowers the community to demystify hardware setups, from building their first NAS to mastering Docker containers.

What to Do When a VM Backup Freezes at the Same Percentage Only in Snapshot Mode - Zima Store Online

A VM backup that freezes at the same percentage only in snapshot mode points to the live-snapshot or guest-freeze path before it points to a fixed bad disk region.

The decisive comparison is snapshot mode versus stop mode using the same VM, destination, and approximate data set. If stop mode repeatedly passes the old percentage, the generic fixed-region diagnosis becomes less likely and the investigation should move to QEMU Guest Agent freeze/thaw, application quiescing, snapshot-capable storage, and live-write behavior. Preserve a working stop-mode backup while testing so troubleshooting never removes the last recoverable copy.

Prove Stop Mode Passes the Snapshot Failure Point

Run one maintenance-window backup in stop mode to the same target used by the failing snapshot job. Record the percentage, disk, throughput, and duration where snapshot mode normally stops.

A Proxmox backup-mode comparison explains that snapshot mode keeps VMs live while stop mode removes many guest-runtime variables.

If stop mode freezes at the same point, return to the generic same-percentage article and source or target storage tests. If stop mode completes, keep the next experiments inside the snapshot-specific path.

Check Whether QEMU Guest Agent Freeze Is the Trigger

Inspect the backup log for guest-fsfreeze-freeze, guest-fsfreeze-thaw, timeouts, or guest-agent communication errors. Compare the timestamp with guest journal or Windows event logs.

A Proxmox Windows backup walkthrough notes that guest agent performs fsfreeze when the agent is enabled.

Do not permanently disable guest consistency features as the first fix. Use a controlled test to determine whether the freeze call itself is the boundary, then repair the guest-agent or filesystem interaction.

Look for Filesystems That Do Not Freeze Cleanly

Inventory mounted filesystems inside the guest, including loop devices, network filesystems, bind-like application mounts, database storage, and unusual control-panel layouts. Note which filesystem is busy when the snapshot begins.

CloudLinux documents a case where fsfreeze can hang complex guests rather than the backup target itself being the cause.

If removing or correcting one problematic guest mount makes snapshot mode pass the percentage, restore consistency protection afterward and document the dependency. Avoid repeated forced resets that can turn a backup problem into guest-filesystem damage.

-15% OFF

ZimaBoard 2 - Mini Home Server for Your Big Idea

Single board computer zimaboard2

Distinguish a Freeze Timeout From a Transfer Stall

Check whether the backup stops before meaningful data transfer begins, immediately after a freeze request, or later during block copying. A fixed percentage near zero can represent a completely different failure stage from a stall halfway through a virtual disk.

A hosting knowledge-base case shows that freeze can block backup indefinitely after the snapshot sequence starts.

If transfer never begins, focus on guest quiescing. If transfer proceeds normally for a long time before stalling, compare live-write load, snapshot layer behavior, and storage latency instead.

Reproduce the Guest Freeze Outside the Full Backup

Where the platform and maintenance policy allow, test the guest-agent freeze and thaw behavior independently or watch the guest closely during a manual snapshot event. Keep a console open and verify that the guest resumes writes after thaw.

A QEMU issue records that QEMU fsfreeze can lock VMs for certain guest filesystem arrangements.

If freeze alone reproduces the hang, repair that guest path before tuning PBS throughput. If freeze and thaw are clean, move downstream into snapshot storage and live-write interaction.

Keep Stop Mode as the Recovery Path Until Snapshot Mode Passes

Do not sacrifice reliable backups while debugging a convenience feature. Schedule stop-mode backups during an acceptable maintenance window until snapshot mode completes several times and a restore test confirms the result.

A Server Fault example shows a real environment using stop mode provides fallback when live snapshot behavior needs separate handling.

The fix is complete when snapshot mode passes the old percentage repeatedly and the guest remains responsive before, during, and after backup. The related ZimaSpace article on generic same-percentage backup freezes remains the correct upstream branch if stop mode also freezes.

Frequently Asked Questions

Is stop mode a permanent replacement for snapshot mode?

It can be a reliable fallback when downtime is acceptable, but snapshot mode is usually preferable when the guest can be quiesced safely and live backups complete consistently.

Should I disable QEMU Guest Agent to make backups work?

Only as a controlled diagnostic test where appropriate. The agent also provides useful shutdown and consistency functions, so the underlying freeze problem should be identified rather than hidden.

Why does the backup stop at the same percentage if the problem is fsfreeze?

Progress percentages can map to repeatable backup phases, not only disk locations. A freeze or snapshot transition reached at the same stage can therefore produce the same visible percentage.