Agent upgrade centrally from B&R server

[perjonsson1960]

Folks,

Is there a way to upgrade the agent packages in the central B&R server, between Veeam upgrades/patches, so that the clients could subsequently be upgraded with the latest version through a rescan? Or do we have to upgrade the agent locally in the client? I have seen that there is a new release of the Linux agent. One of our physical Linux (Ubuntu) machines have failed to backup since we upgraded the core in that machine from 5.3 to 5.4.

Failed to take volume snapshot
Failed to perform managed backup
Child execution has failed. Exit code: [1].
Failed to load module [veeamsnap] with parameters [zerosnapdata=1 debuglogging=0 snapstore_block_size_pow=14 change_tracking_block_size_pow=18 logdir=/var/log/veeam fixflags=0].

PJ

[PTide]

Hi,

Yes, it is possible. Please see this page.

However, VBR does not support agents that have been released later than VBR. E.g. v9 does not support v4, v10 does not support v5 and v4.0.1.
It might work though, but you need to keep in mind that some fixes that are included in v4.0.1 require a patch to be applied on VBR as well.

That is, it is advisable to upgrade VBR to v10a first so it can upgrade all agents automatically.

Thanks!

[perjonsson1960]

We are running B&R v10.0.0.4461 P2 (Cumulative Patch 2). Is there a later patch?
The client in question currently runs Linux agent v4.0.0.1961.
Are you saying that our version of B&R does not support v4.0.1 of the Linux agent?

PJ

[PTide]

Yes, for 4.0.1 you need 10.0.1.4854.

Thanks!

[perjonsson1960]

Thanks!
I have now upgraded to 10.0.1.4854.
And I have managed to upgrade the Linux agent to v4.0.1.2365 in our two physical servers. However, now one of them complains about some certificate, and the B&R server is unable to connect via ssh, and that Linux server has disappeared from the protection group... I guess that means no backup...

[PTide]

Hi,

Kindly open a support case to resolve the issue and post the case ID in this thread.

Thanks!

[perjonsson1960]

We have managed to get the rescan to work. Apparently there were some additional sudo commands that needed to be allowed. And the agent is now 4.0.1.2365, as I mentioned before. However, now this particular Linux machine is being completely ignored by the backup job. It doesn't even show up in the "Building hosts list" at the beginning of the job. And I even created a new backup job containing only that machine, and it was completely ignored in that job, as well. The job just says "Nothing to process" and stops. This is so weird, and I am completely in the dark...

PJ

[PTide]

If the Agent has been upgraded to 4.0.1 and the rescan goes well, I would suggest opening a support ticket. Please post you case ID in this thread.

Thanks!

[perjonsson1960]

Case ID: #04333292

[perjonsson1960]

The problem was due to conflicting machine entries in the Veeam database. We deleted the machine from the backup job and from the protection group, deleted the corresponding machine entries in the database, created a new temporary protection group with only the machine in question, created a new temporary backup job and added the protection group (not the individual computer) to the backup job, and ran the job. It worked.

After that, I deleted the temporary backup job and the temporary protection group, added the machine to the original protection group, and then added the machine to the original backup job (as an individual computer), and that worked, as well. Problem solved by the excellent Veeam support!