Health Check fails for Immediate Copy Job & Agent Backups

[flomp]

I have the following setup
- VM-Backups daily to local B&R-Repo (not really relevant here)
- Agent-based backups for Windows 11 Pro machines to local B&R-Repo
- Immediate Backup Copy Job to ROTATED USB-storage. Disks a rotated daily (5 disks, Monday to Friday)
- Starts in the evening at 20:00 (takes one our at most for all machines)
- Health Check enabled for the backup copy job (Monday to Friday)
- Scheduled for 1:00 in the morning
- The idea was to have everything copied over and then check whether the data is ok.

Following Problem
- If a Windows 11 Pro machine is not running during the day, it won't create a backup
- So no backup will be copied to the USB storage
- But, the health check expects the backup from the day before to be there. This is obviously not the case because the disk has been rotated and the email-report shows an error.

What solutions do I have?
- Change to periodic copy?
- Anything else?

Is this a known problem? It's unfortunate that an error mail is created even though there is not really a problem.

[Dima P.]

Hello flomp,

Sorry for the delay and thank you for your post!

But, the health check expects the backup from the day before to be there. This is obviously not the case because the disk has been rotated and the email-report shows an error.

Healthcheck should not generate any warning or errors since no new restore point was created. Do you have a case ID for it? Thank you!

[flomp]

Hey, yes it is #07448582.

And you are right. In the meantime I switched to periodic copying and the error remained.

After checking with support, it turned out the reason was that the machine has not been rebooted since a few weeks. After we did the reboot, the error message disappeared. However it started to appear on another machine and disappeared there again after rebooting.

It would be good, if the error message could be improved.

At the moment it looks good and I am monitoring the situation. I want to try to switch back to immediate copying in the next few days.

[flomp]

Healthcheck should not generate any warning or errors since no new restore point was created. Do you have a case ID for it? Thank you!

I just got errors again and updated the case.

[flomp]

After investigating some more and also having a Webex call with support, it turned out that Veeam does not keep track on which rotated disks which backup copies have been stored and so it expects the last copied VIB to be there - EVEN IF THE DISK HAS BEEN ROTATED.

I can't understand why this has been implemented so naively. This renders the health check useless, as I have to check everyday if there was a real error or not. It should be easy for Veeam to
1. either keep track on which disks which VBKs/VIBs have been stored and just check those
2. or at least just take the VBKs/VIBs it finds if there has not been a backup copy for the machine since the last disk rotation

However, this explanation differs from what you said:

Healthcheck should not generate any warning or errors since no new restore point was created. Do you have a case ID for it? Thank you!

Can you please confirm that what I found out with support is correct?

The workaround which I was told was to set the "Delete Backups when drive has been changed" setting for the rotated drives-repository. But this would mean that I only have one backup on each disk!

The setup which I ran the last ~2 weeks is a bit different from the one in the first post, so for completeness, here are the current settings:
- There are 5 USB disks, which are rotated daily (Monday to Friday)
- There is a backup policy for the workstation machines to create a backup to a local repo daily at 12:30
- There is a copy job for copying everything to the rotated USB disks (periodic, 23:00)
- Health check for the copy job is scheduled for 2:00 in the night

Everything works fine if all 5 workstations created a backup during lunch. Those backups will be copied to the rotated USB drive and then the healtch check will run.
(But why doesn't it complain that it can't find the VIBs of the other weekdays??)
But if a workstation did not run and therefore did not create a backup during lunch, there is obviously nothing to be copied to the USB disk. However the backup copy job will still run without error and copy the other backups. (This is what I expected.) But then, for the machines which did not create a backup during the day, the health check will fail with the message:

Code: Select all

[...] Failed File [f:\backup\COPY-JOB-NAME\BACKUP-JOB-NAME - COMPUTERNAME\BACKUP-JOB-NAME - COMPUTERNAMED2024-09-26T230327_B9E0.vib] does not exist. Failed to verify storage [...] metadata. Agent failed to process method {Stg.CheckMetadataCorrupt}.

If this is really the way which Veeam currently should behave, I strongly ask for a feature request (tbh: I would call it bugfix, because it is not really the way someone expects Veeam to behave) that implements one of the two variants outlined above.