backup owner check failed from the same cluster ?

[hmelgers]

Hi,
I'm running a 3 node pve cluster that use the same pbs. After adding a first vm to a second node and trying to backup i get this error:

301: 2021-09-01 20:00:02 ERROR: VM 301 qmp command 'backup' failed - backup connect failed: command error: backup owner check failed (evuser@pbs != biuser@pbs)

Backups from the first node run without problems to the same pbs.
This is a cluster so i assume there's only one place storing the pbs credentials. It seems it is using the right credentials but somehow pbs expects another user ?

I'm confused. Anyone knowing what is going on here ? pve version 6.4, pbs version 1

 

[dylanw]

Hi,

The nodes should all be using the same storage configuration in the background. Is it possible that you have a totally different VM301 backup already on the PBS datastore, owned by user 'biuser@pbs', which could be interfering?

 

[hmelgers]

You're absolutely right Dylan, that must be the cause. I do have another 301 from that different user, wasn't aware that id was all leading.
It would be nice to have some prefix digits for each vm on a pve to be set to avoid these kind of circumstances.

Thanks for your help!

 

[dylanw]

Happy I could help
Although I can't confirm any specific time frame, allowing for these scenarios is a part of our PBS roadmap [1].

[1] https://pbs.proxmox.com/wiki/index.php/Roadmap