Backups and restore

A backup is a complete, self-contained copy of a virtual machine that can be restored onto any node. This guide sets up a scheduled backup job, then restores from it.

Snapshots are not backups

A snapshot records a point in time on the same storage as the machine itself. If that storage fails, the snapshot is lost with the machine. Use snapshots for a quick, reversible checkpoint before a risky change; use backups for anything you would need after losing a disk or a node.

Before you begin

• Decide where backups will be written. Any storage with Backup in its content types can be a target, including a dedicated VM2Cloud Backup Server if you run one.
• Decide how many copies to keep. Retention is set per job.
• Backups are read-heavy on the source storage. Schedule them outside your busiest hours.

1. Check the backup storage

1. Open Datacenter → Storage.
2. Select the storage you intend to use and click Edit.
3. Confirm Backup appears in Content. If it does not, add it and click OK.

Note

local-zfs holds virtual machine disks and does not accept backups. local is enough to try your first backup. For backups you rely on, add dedicated backup storage on different hardware — see Storage.

2. Create a scheduled backup job

Backup jobs are configured for the whole datacenter, not per node.

1. Open Datacenter → Backup.
2. Click Add.
3. Set Storage to your backup target.
4. Set Schedule to when the job should run — for example 02:00 for every night, or sat 02:00 for Saturdays only.
5. Under Selection mode, choose which guests are included:
   • Include selected VMs — tick the machines you want.
   • All — every guest on the cluster, including ones created later.
   • Exclude selected VMs — everything except the ones you tick.
   • Pool based — every guest in a resource pool; adding a guest to the pool adds it to the job.
6. Set Mode — see the table below.
7. Open the Retention tab and set how many copies to keep.
8. Click Create.

The job appears in the list and runs at the scheduled time. Select it and click Run now to test it immediately rather than waiting.

Backup modes

Mode	What happens	When to use it
Snapshot	The machine keeps running. The disk is captured at a point in time.	The default. Use it unless you have a specific reason not to.
Suspend	The machine is paused for the copy, then resumed.	Rarely needed; snapshot mode is usually better.
Stop	The machine is shut down, copied, then started again.	When you need a guaranteed-consistent copy and can accept downtime.

Make snapshot backups application-consistent

In snapshot mode the guest keeps writing during the copy, so the result is crash-consistent — equivalent to pulling the power. Installing the guest agent inside the virtual machine lets VM2Cloud VE briefly freeze the guest filesystem so the copy is clean. Enable it under the machine’s Options → QEMU Guest Agent, and install the matching agent package inside the guest operating system.

3. Confirm the backup ran

1. Open Datacenter → Backup and select the job, or look at the Tasks list at the bottom of the screen.
2. A successful run finishes with TASK OK.
3. Open your backup storage in the Server View, then Backups, and confirm the archive is listed with the expected date and size.

Test a restore before you rely on it

A backup job reporting success only proves a file was written. Restore at least one machine to a new VM ID and boot it, so you know the archives are usable before you need them.

4. Restore a virtual machine

To restore alongside the original, as a new machine:

1. In the Server View, select the storage holding the backups.
2. Click Backups.
3. Select the archive you want and click Restore.
4. Set Storage to where the restored disks should be written.
5. Set VM ID to a free ID, so both machines can exist at once. This dialog does not accept an ID that is already in use.
6. Click Restore.

To restore over the original machine:

1. Shut down the machine. A running machine cannot be restored over; the restore task fails and leaves it as it was.
2. In the Server View, select the machine itself and open Backup.
3. Select the archive you want and click Restore.
4. Set Storage to where the restored disks should be written. The VM ID is fixed to this machine.
5. Click Restore and confirm the warning that the machine’s current data will be erased.

Restoring over an existing machine destroys it

Restoring to a VM ID that is already in use replaces that machine and its disks entirely. If you are not certain, restore to a new ID instead and compare the two before removing anything.

Watch the task window until it reports TASK OK, then start the restored machine and confirm its data.

Snapshots — a quick checkpoint

Use a snapshot before an upgrade or a configuration change you might want to undo.

1. Select the virtual machine.
2. Click Snapshots, then Take Snapshot.
3. Give it a name and, optionally, a description.
4. Tick Include RAM if you want the machine’s running state captured as well as its disk.
5. Click Take Snapshot.

To go back, select the snapshot and click Rollback. To free the space again, select it and click Remove.

Note

Snapshots accumulate storage as the machine’s disk diverges from them. Remove ones you no longer need rather than keeping them indefinitely.

If something goes wrong

What you see	What to do
The storage is not offered as a backup target	Open Datacenter → Storage, edit the storage, and add Backup to its Content.
The job does not run at the scheduled time	Re-check the Schedule field, and confirm the node’s clock is correct — see Fixing “Authentication failed! (401)”.
The backup fails part-way with no space left	Check free space on the backup target, and lower Retention so older archives are pruned.
Restoring over a machine fails because it is running	Shut the machine down, then restore again.
The restored machine will not start	Confirm you restored to storage the node can reach, and check the Tasks list for the restore result.
The restored machine has no network	The restored configuration keeps its original bridge and VLAN. Check them under Hardware against the node you restored onto.

Still stuck? Contact VM2Cloud support.