2 synthetic full backups will be placed into the media set

[tohhmas]

I’m looking for some clarity an issue we’re seeing with our monthly tape backup.
As per the title, we are getting duplicated backups in the tape media set. This has caused lasts weekend’s attempt to fail and seems to be happening this weekend too.

08/11/2018 15:14:23 :: No restore point for 02/11/2018 is available for virtual full backup, using previous restore point for 31/10/2018
08/11/2018 15:14:24 :: 2 synthetic full backups will be placed into the media set
09/11/2018 01:43:17 :: No restore point for 09/11/2018 is available for virtual full backup, using previous restore point for 07/11/2018

The above lines in the job output show that it starts working on the second full backup set once the first (older) set has been written to tape. There should be no reason why we write the older restore point to tape.

This has only started recently, but basically causes the whole process to fail as we run out of available tapes as we’re doubling the data that’s getting written. We disable all source jobs whilst the tape job is running to stop retries when the source backup files change (although ideally we would like not to be forced to do this as the tape job can take upwards of 100 hours to complete)

Can anyone think of a setting that would cause this to happen? My team members are saying that nothing's been changed..

Support case ID: 03280018

[Dima P.]

tohhmas,

According to the provided log portion there was no restore points for the given dates, so the closest available restore point to the given date was used to create a synthetic full backup. Please keep working with our support team, it should be possible to provide more detailed response to your questions based on application debug logs. Cheers!

[tohhmas]

Yes, what you're describing is intended behavior and not what my ticket is about.
The issue is that 2 full backups were being placed in the media set.

For those who stumble across this thread in the future:
Our support engineer said this was unusual behavior and most likely a glitch. We have since restarted both the Veeam management server and Veeam repositories and the issue has subsided for now.

[lyapkost]

Hi.

There should be no reason why we write the older restore point to tape.

There is a reason: When the problem is fixed, the GFS job will archive one most recent missed full backup for each media set. It seems that for some reason there's no weekly backup on tapes for date 02/11/2018 so when the job starts it processes two backups: for 02/11/2018 (last week) and for 09/11/2018 (current week).

If you do not want GFS tape job to process old backups just disable the tape job and enable it again, this will fix your issue.

[tohhmas]

This is not a GFS tape job, it just uses a simple tape Media Pool.

The only use of GFS we have in our Veeam installation is in our Backup Copy jobs and these have no impact on the tape job.

[Dima P.]

tohhmas,

Thanks for the heads up. Can you please clarify the synthesized full backup creation schedule (Backup to tape job > Full Backup > Schedule button)? Cheers!

[tohhmas]

The issue reoccurred again on the 3rd attempt since the restart. (The first two tests didn't have the issue).

I've since re-created the tape job using identical settings and for the first run it seems to have resolved the issue again, that's not to say it won't reoccur again though!

Will continue to monitor.

[tohhmas]

tohhmas,

Thanks for the heads up. Can you please clarify the synthesized full backup creation schedule (Backup to tape job > Full Backup > Schedule button)? Cheers!

Here's the schedule,

[lyapkost]

Checking the checkbox Process latest full backup chain only should help in your case then.

[tohhmas]

That's already enabled I'm afraid.

[Dima P.]

tohhmas,

Thanks for the update, settings look good. Please keep monitoring your tape backups and if you see the same issue again raise a new case. Do not forget to update this thread with new case ID. Cheers!