How to copy successfully from an existing SOBR

[mmasching]

Hi,

we are trying to add a capacity tier with immediate copy to an existing 120TB SOBR.
However, it is creating a lot of subjobs due to the number of existing backups/restore points to be copied (8000+).
This overloads the backup server.

Even by doubling the host memory it is not sufficent.

Is there a chance to select backup jains to be copied? Currently we only have 1 full and 14 incs as restore points for the jobs. So it is not possible to "move" older backups first.

regards
Marcus

[Gostev]

The number of restore points should not cause any issues, because they are not processed all simultaneously, but rather according to available backup infrastructure resources (and as the lowest priority task). Most likely, you have concurrent tasks value set to high on your repositories, and your available RAM resources don't match that value based on our system requirements. You should let our support take a look at your setup, they will suggest the required changes. Thanks!

[mmasching]

Hi,
I checked the concurrent tasks on the proxies and they were slightly too high - adjusted.
However, the Repo is a flashbased NAS and I can't see any saturation there.

The subjobs fail all with the same error:

Code: Select all

14.07.2020 04:04:22 :: Connection to 'Veeam Backup Service' is not available, host 'XXXXX', port '9392'  
14.07.2020 03:30:29 :: Processing BJ-XXXXXX Error: Remoting Channel Sink UriNotPublished. RequestUri is '/a4c5decf_4ecf_4bbb_b2d6_218b1959944b/sa1kzlnysoi2kzfcabjy9qqg_158237.rem'  

BS is upgraded to 96GB RAM.

Gonna raise a case like suggested.

Thanks
Marcus

[veremin]

Based on the error, it doesn't seem to be related to object storage specifically, but let's give a support engineer a chance to figure this out.

Kindly, post your support ticket Id, once you have it.

Thanks!

[mmasching]

Case #04282048 — Offload backup to Cloud object Storage failed

[veremin]

Thanks, Marcus, now let's wait and see what support engineer suggests after log investigation. Thanks!

[mmasching]

Hi Vladimir,

repsonse of the support was not satisfactory. Would you please have a look if possible?

Could it be, that SOBR Offloads, create a dedicated jobs for each VM to be offloaded? and this will exceed the ressources of the Backup Server?

Although there are currently only 51 jobs running in the console, we have a process count of nearly 700 on the BS


regards
Marcus

[Gostev]

I checked and the case is in good hands now, with engineers looking at the right things, so let's give them sufficient time to troubleshoot. The only thing I'm not entirely happy about is an unjustified ask to double RAM. Although considering how many times it fixes all issues, may be it was a good first step

[mmasching]

Hi,

followed the instructions of the case and rebooted the DB-Server. Additionally increased Memory for the db from 40GB to 60GB.
Unfortunately behavior is still the same. Lots of failing SOBR Offload subjobs due to missing connection to Backup Server.
reported back on the case, too.