LXCs fail to stop for and restart after backup randomly

whostolemyname

New Member
Mar 1, 2026
14
2
3
Hello there,
I am running a PBS inside a LXC on Proxmox VE (because at this time I do not have dedicated hardware for PBS). The backups however are being stored on a disk separate from the root fs disk. Everything works fine, except for the fact, that every now and then random lxcs fail to restart after backup or fail to stop for the backup. This is somewhat crucial, when it happens with the lxc where a pihole-server is running, because a not running pihole means no running internet. The error doesn't happen on a regular basis, so I cannot reproduce it for troubleshooting. Also, since it doesn't happen on a regular basis, I would assume that the setting for the backup are not the culprit.
Has anyone in the forum encountered a similar behaviour of PBS?
 
I'd start with the failed backup task on the PVE node that runs the affected container. In the task output, find that CT's section and note the actual backup mode and the first error. In stop mode the container is stopped for the backup. In snapshot or suspend mode, the task can suspend and resume it, so the point where it fails matters. PVE backup modes

If the log gets as far as trying to start the CT again, check its service journal on that same PVE node. For CT 123, and assuming the failure was recent:

Bash:
journalctl -u pve-container@123.service --since "2 hours ago"

Replace 123 and adjust the time window to cover the failed run. If you catch it again, the first error and a few lines around it from the backup task would help show whether it got stuck stopping the CT, writing the backup, or bringing it back up. PVE container service
 
Hi,
the task log starts as follows:

Rich (BB code):
INFO: starting new backup job: vzdump 201 202 204 205 203 206 --quiet 1 --notes-template '{{guestname}}' --mode stop --fleecing 0 --prune-backups 'keep-daily=3,keep-monthly=2,keep-weekly=2,keep-yearly=1' --notification-mode notification-system --storage PBS_PVE_Backups
INFO: Starting Backup of VM 201 (lxc)
INFO: Backup started at 2026-11-18 00:00:04
INFO: status = running
INFO: backup mode: stop
INFO: ionice priority: 7
INFO: CT Name: LXC12-PiHole
INFO: including mount point rootfs ('/') in backup
INFO: stopping virtual guest
command 'lxc-stop -n 201 --nokill --timeout 600' failed: exit code 1
ERROR: Backup of VM 201 failed - container did not stop
INFO: Failed at 2026-11-18 00:20:09
INFO: Starting Backup of VM 202 (lxc)

after the last entry the backup ran a couple of other containers. The backup job is running daily, the mode ist "stop". As mentioned before, the failure doesn't happen regularly and it doesn't always affect the same lxc.

Journalctl states:

journalctl.png

I have started the container at 10:13 manually.