PBS backup email notification incomplete

[keson]

Dear community,
not that this is a problem that would affect the Proxmox environment reliability, but more an observation I made in past few weeks/months.
Running Proxmox 8.2.7 and PBS 3.2.7 with several backup jobs running daily to backup all VMs and CTs to external storage, email notifications set as default(auto), sent always.
There is a set off 9 VMs/CTs across 2 Nodes and the backup runs as expected.

The only weird thing is the notification email content. Although the Backup job backs up ALL expected 9 VMs/CTs, the notification email contains only 7 VMs/CTs from one node.

So the log email looks pretty much like this:

Details​

VMID	Name	Status	Time	Size	Filename
111	dc02.xxx	ok	3min 31s	32.001 GiB	vm/111/2024-10-30T21:35:03Z
112	dc03.xxx	ok	3min 26s	32.001 GiB	vm/112/2024-10-30T21:38:34Z
210	ts02.xxx	ok	29min 42s	256.001 GiB	vm/210/2024-10-30T21:42:00Z
500	rport.xxx	ok	35s	32.001 GiB	vm/500/2024-10-30T22:11:42Z
501	docker.xxx	ok	11s	32.001 GiB	vm/501/2024-10-30T22:12:17Z
2810	ms.nxxx	ok	20s	4.006 GiB	ct/2810/2024-10-30T22:12:28Z
2850	unifi.xxx	ok	27s	3.516 GiB	ct/2850/2024-10-30T22:12:48Z

Total running time: 38min 12s
Total size: 391.525 GiB

Logs​

vzdump 210 2850 111 112 500 501 700 110 2810 --fleecing 0 --storage NAS_NFS_PBS_Backup --quiet 1 --mode snapshot --mailnotification always --notes-template '{{node}}, {{vmid}}, {{guestname}}' --mailto admin@mymail.com

As you can see the vzdump covers all 9 targets, but the summary details table shows only 7 of them, the ID 110 and 700 are left out - they both sit on pve01, while the other 7 sit on pve02.

Also the details log in the email body only covers the 7 targets from the pve02.

Anyone have seen similar behavior?

As I said, the backups are done, nothing is left out, only the notification is incomplete

 

[fiona]

Hi,
each node will send its own notification. Please check the logs and configuration of the other node.

vzdump 210 2850 111 112 500 501 700 110 2810 --fleecing 0 --storage NAS_NFS_PBS_Backup --quiet 1 --mode snapshot --mailnotification always --notes-template '{{node}}, {{vmid}}, {{guestname}}' --mailto admin@mymail.com

As you can see the vzdump covers all 9 targets, but the summary details table shows only 7 of them, the ID 110 and 700 are left out - they both sit on pve01, while the other 7 sit on pve02.

If you read the log you should see that 700 and 110 are skipped as external. There is a second vzdump invocation on the other node.

 

[keson]

Hi,
each node will send its own notification. Please check the logs and configuration of the other node.


If you read the log you should see that 700 and 110 are skipped as external. There is a second vzdump invocation on the other node.

You are absolutely right! I did not look at it from this angle - my bad - I did not expect the log file ON THE SERVER has some extra lines compared to a logic set by email - namely the following:

INFO: starting new backup job: vzdump 210 2850 111 112 500 501 700 110 2810 --fleecing 0 --storage NAS_NFS_PBS_Backup --quiet 1 --mode snapshot --mailnotification always --notes-template '{{node}}, {{vmid}}, {{guestname}}' --mailto admin@myemail.com
INFO: skip external VMs: 110, 700

So I have to investigate, why the second server sends emails OK and why the emails from the first server are lost.

All clear, hopefully toning both emails will arrive