job-id: invalid format - invalid configuration ID

[automaton]

Ever since I updated to 8.2.5, I've been getting the following error in my log when my backup jobs are supposed to run once a week:

Sep 28 01:00:09 pve pvescheduler[117971]: 400 Parameter verification failed.
job-id: invalid format - invalid configuration ID 'b61e26046051f94aef2862d364520ad621e35710:1'
Sep 28 01:00:09 pve pvescheduler[117971]: jobs: encountered object '400 Parameter verification failed.
job-id: invalid format - invalid configuration ID 'b61e26046051f94aef2862d364520ad621e35710:1'


', but neither allow_blessed, convert_blessed nor allow_tags settings are enabled (or TO_JSON/FREEZE method missing) at /usr/share/perl5/PVE/Jobs.pm line 228.

I've only ever used the UI to create backup jobs on the data center, never the command line, so I don't understand how this could happen, or how I can fix it.

If I run the backup job manually it works fine with this log:

Sep 29 19:45:22 pve pvedaemon[1235232]: INFO: starting new backup job: vzdump --quiet 1 --node pve --exclude 110,109 --all 1 --storage pve-backups --mode snapshot --fleecing 0 --mailnotification always --compress zstd
Sep 29 19:45:22 pve pvedaemon[1235232]: INFO: Starting Backup of VM 100 (lxc)
Sep 29 19:45:22 pve dmeventd[506]: No longer monitoring thin pool pve-data-tpool.
Sep 29 19:45:22 pve lvm[506]: Monitoring thin pool pve-data-tpool.
Sep 29 19:45:23 pve kernel: EXT4-fs (dm-18): write access unavailable, skipping orphan cleanup
Sep 29 19:45:23 pve kernel: EXT4-fs (dm-18): mounted filesystem ec7e7790-e54d-4856-96a3-3ad5a3d3d3ea ro without journal. Quota mode: none.
Sep 29 19:46:10 pve kernel: EXT4-fs (dm-18): unmounting filesystem ec7e7790-e54d-4856-96a3-3ad5a3d3d3ea.
Sep 29 19:46:10 pve pvedaemon[1235232]: INFO: Finished Backup of VM 100 (00:00:48)
Sep 29 19:46:10 pve pvedaemon[1235232]: INFO: Starting Backup of VM 101 (lxc)

Any ideas what's going on here or how to fix it? Thanks.

 

[dcsapak]

this was a bug: https://bugzilla.proxmox.com/show_bug.cgi?id=5731
and is already fixed: https://lore.proxmox.com/pve-devel/20240920090506.231780-2-f.gruenbichler@proxmox.com/
since pve-manager 8.2.6 which should already be in the pve-enterprise repository

 

[jazzl0ver]

this was a bug: https://bugzilla.proxmox.com/show_bug.cgi?id=5731
and is already fixed: https://lore.proxmox.com/pve-devel/20240920090506.231780-2-f.gruenbichler@proxmox.com/
since pve-manager 8.2.6 which should already be in the pve-enterprise repository

just discovered that my backups didn't run since September..

here is the logwatch config that should catch and report proxmox errors (if any):

cat /etc/logwatch/scripts/services/proxmox
#!/bin/bash

grep -i " pve" | egrep -i "(error|critical|failed)"

cat /etc/logwatch/conf/services/proxmox.conf
Title = "Proxmox Logs"
LogFile = syslog

cat /etc/logwatch/conf/logwatch.conf
Service = "proxmox"

one may test it with logwatch --service proxmox --range today

 

[Maximiliano]

just discovered that my backups didn't run since September..

Please check your notification settings, you should have received an email for each failed backup. The notification panel in the web UI has a "Test" button you can use to check whether notifications work.

 

[jazzl0ver]

That's the point - I do have backup failure notifications set, but they weren't sent. Looks like it's b/c the error happened before that actual backup had started.

 

[Maximiliano]

> I do have backup failure notifications set, but they weren't sent

Sounds like notifications were not correctly setup. Make sure that the backup job uses "notification system" instead of a specified email, and that the root user has an email set.

 

[jazzl0ver]

> I do have backup failure notifications set, but they weren't sent

Sounds like notifications were not correctly setup. Make sure that the backup job uses "notification system" instead of a specified email, and that the root user has an email set.

Thank you very much! Indeed, the Notifications were not set up and used the default root user (root@pam) without a proper email set.

However, I still think it might be useful to have a separate logwatch rule to catch Proxmox issues in the system logs. Would be great if you managed to add it to the logwatch upstream (I can't register on Sourceforge for some reason to push a patch).