Backups not timing out

[jefff]

We run a single nightly backup job to back up all of our VMs. At least one node every night will get stuck on a VM for hours and will sit there indefinitely until we cancel the job (and all other VMs that weren't backed up yet will not be done since we cancelled it). I tried to add the following to our vzdump.conf but when I run the jobs, the first thing it says is "ignore config line:" and ignores my changes.

timeout 10800
lockwait 3600
stopwait 600

We created the backup job via the UI. I can't figure out how to fix this problem. Any help is appreciated. We're still very new to Proxmox so please excuse me if I ask a lot of questions.

 

[jklocker]

/etc/vzdump.conf uses colon-separated syntax - option: value. See https://pve.proxmox.com/wiki/Backup_and_Restore#vzdump_configuration.
Values for lockwait and stopwait are in minutes, not seconds. Also timeout isn't a valid option.

 

[jefff]

/etc/vzdump.conf uses colon-separated syntax - option: value. See https://pve.proxmox.com/wiki/Backup_and_Restore#vzdump_configuration.
Values for lockwait and stopwait are in minutes, not seconds. Also timeout isn't a valid option.

Thanks for this. That's what I get for trusting AI.

I fixed the syntax. It no longer ignores the values, but it still doesn't appear to be using them. Here's the first line of the backup job when it's run:
INFO: starting new backup job: vzdump --prune-backups 'keep-daily=65' --notes-template '{{guestname}}' --node pve7 --fleecing 0 --all 1 --mode snapshot --storage pbs

Also, can you speak to the issues I'm having? Do you think the lockwait and stopwait would help remedy the VMs getting locked up while the backup sits there trying to complete? I want to make it so if a backup of a vm takes longer than an hour, it skips it. But I can't figure out how to do so.

 

[jklocker]

The config is still applied. The UI job only shows the flags it passes explicitly, and anything not passed falls back to /etc/vzdump.conf. So your lockwait/stopwait values aren't being ignored just because they're absent from that INFO line.
That said, lockwait and stopwait won't terminate the backup after a certain time..

You should look into why the backup hangs. Is it always the same VM? What do the logs show?