Problem with backup restore

[Tom78]

Hi,
I'm new to proxmox, so perhaps its a common problem and I'm just doing something wrong.
I made a backup of a VM before reinstalling proxmox from ground up (new CPU, new SSDs, new NICs, so I reinstalled). There were no problems during backup and from the 5 VMs I made a backup 4 are restoring fine, the biggest one does not want to restore.
This is the end of the backup log, no errors, all was fine :

2024-04-12 23:12:25 INFO: 98% (593.8 GiB of 604.0 GiB) in 37m 55s, read: 1.6 GiB/s, write: 11.0 KiB/s
2024-04-12 23:12:28 INFO: 99% (598.1 GiB of 604.0 GiB) in 37m 58s, read: 1.4 GiB/s, write: 4.0 KiB/s
2024-04-12 23:12:31 INFO: 100% (604.0 GiB of 604.0 GiB) in 38m 1s, read: 1.9 GiB/s, write: 5.3 KiB/s
2024-04-12 23:12:31 INFO: backup is sparse: 466.93 GiB (77%) total zero data
2024-04-12 23:12:31 INFO: transferred 603.99 GiB in 2281 seconds (271.1 MiB/s)
2024-04-12 23:12:47 INFO: stopping kvm after backup task
2024-04-12 23:12:52 INFO: archive file size: 0KB
2024-04-12 23:12:52 INFO: adding notes to backup
2024-04-12 23:12:52 INFO: Finished Backup of VM 100 (00:38:23)

Now if I want to restore this one, at 87% I get errors and can't restore:

progress 87% (read 564218953728 bytes, duration 1264 sec)
_12-22_34_29.vma.zst : Read error (39) : premature end
progress 88% (read 570704199680 bytes, duration 1265 sec)
progress 89% (read 577189511168 bytes, duration 1266 sec)
progress 90% (read 583674757120 bytes, duration 1266 sec)
progress 91% (read 590160068608 bytes, duration 1266 sec)
progress 92% (read 596645314560 bytes, duration 1266 sec)
vma: restore failed - detected missing cluster 9099367 for stream drive-sata0
/bin/bash: line 1: 3645 Exit 1 zstd -q -d -c /mnt/pve/Win10/dump/vzdump-qemu-100-2024_04_12-22_34_29.vma.zst
3646 Trace/breakpoint trap | vma extract -v -r /var/tmp/vzdumptmp3634.fifo - /var/tmp/vzdumptmp3634
Logical volume "vm-102-disk-0" successfully removed.
temporary volume 'local-lvm:vm-102-disk-0' sucessfuly removed
Logical volume "vm-102-disk-1" successfully removed.
temporary volume 'local-lvm:vm-102-disk-1' sucessfuly removed
no lock found trying to remove 'create' lock
error before or during data restore, some or all disks were not completely restored. VM 102 state is NOT cleaned up.
TASK ERROR: command 'set -o pipefail && zstd -q -d -c /mnt/pve/Win10/dump/vzdump-qemu-100-2024_04_12-22_34_29.vma.zst | vma extract -v -r /var/tmp/vzdumptmp3634.fifo - /var/tmp/vzdumptmp3634' failed: exit code 133

Is there anything I can do ? How is it possible, that the backup process had no errors and now I can't restore ? Hope someone has an idea.

 

[Tom78]

Nobody any ideas ? Perhaps there is a tool to try to fix the corrupted file ? It would be very sad, if I loose the VM.

 

[Tom78]

I don't know. To be 100% on the safe side, I made 2 backups of every VM. Now I bit to bit compared the two backups of the corrupt big VM and both are 100% identical. There also were no problems during both backup processes. How is this possible ? Both backups are the same and both had no errors, but during restore both have the same error at 87%. I hope there is someone with an idea, the data is very important for me (a lot of old photos) and to be honest, i don't know what to do now. As I said I made the backups twice, because the data was so important for me and now I have two corrupt files, it seems.

 

[intrax]

That's a sad story!
Have you tried to restore the backup (vzdump file) on another linux machine ?
Try to verify the tar file first ( tar -tzf my_tar.tar.gz >/dev/null ) to see if it's corrupted...

 

[Tom78]

It seems to be corrupt, but I dont know why it happened twice without any error message. I already tried on another machine, same problem. Will try to verify the tar file, didnt think about it, thank you for the advice.

 

[intrax]

On todays update install I noticed an update to pve-backup-restore. Have you tried it?

 

[Tom78]

Didnt see it, thank you, will try it later.