Relocate Single Active Full on SOBR

[nunciate]

I have searched around but have not found the exact process for this so I figured I would post and ask.

I have a Scale-Out Repository with 5 drives in it. 2 of the drives are getting pretty full while one of them has over 10Tb free.
I want to selectively move some full backups from the almost full extents to the one with 10Tb free.

What is the process for this? The scale out is set for Performance Policy with All extents allowing all backup types and I am using Per-VM Backups.
So on my backup jobs will have the VBKs on one extent and the VIBs on another.

I don't want to evacuate the entire extent that is getting full. I just want to move some things around to balance all extents out a bit more.
Can I simply move the folder with the VBK from one to the other extent and then do a Re-scan Repository or do I need to remove the backup from the configuration and re-map it within the job?

Can someone provide the exact steps for this. Might I suggest this become a KB article, FAQ or something we can search for on the main support site?

[Vitaliy S.]

Can I simply move the folder with the VBK from one to the other extent and then do a Re-scan Repository

These steps should be sufficient according to our QC team.

[nunciate]

Great I will give it a try.

[Vitaliy S.]

Please post back the results for our future readers. Thanks!

[nunciate]

I can confirm that this did, in fact, work as expected.

I had a 3Tb VBK on the full extent and the same job had about 400Gb of VRB files on the destination extent.
I disabled the backup job.
I renamed the folder with the VBK temporarily and robocopied it over to the destination extent.
Once completed I deleted the folder off of the full extent
I then robocopied the folder with the VRB files from the destination extent over to the other extent just so I can keep them separated like they were.
I renamed the folder with the VBK file back to it's original name and performed a Re-Scan on the Scale-Out Repository.
The job stated that changes were detected and updated the job. I then pulled the properties of the disk backups for the job and all of the files showed the new locations.
I ran an incremental backup which ran successfully and copied the incremental data to the correct/new location.

[Vitaliy S.]

Thanks for the update!

[fraack]

Hi Folks,

last week we installed a second repository for one of my customers. Both repsoitories are standalone servers with local disks. We created a SOBR to extend the existing capicity. Since we ran every backup job several months before implemeting the new repository all backup chains are created at our first extend. SOBR runs with per VM backup chain and data locality policy. Is there any way or simple workaround to rebalance some of the existing backup chains to the new extend. I do not want to run into a full repsoitory and aditionally take advantage of the extend performance.

Does any of you have some experience with this topic? I thought to evacuate some chains manually and map the jobs to the new repo. Or to set the existing extend to maintanence mode, evacuate the data and interrupt this procedure after a couple of hours. But may some of you know a smoother way.

Thank you

Florian

[DaStivi]

Hallo Florian,
i would simply try to move the files with windows explorer and do a rescan on the backup repositories and than on the sobr repo itself..
when using refs than its something more difficult...

lg

[foggy]

Stephan is spot on, moving files around with subsequent rescan should work.

[rogerdu]

Is it possible to re-balance the extent usage on a Scale Out Repository? of 3 extents, I have one that only has 1.5TB, one that has .8TB and one that has 9.5 TB free...

Suggestions?

[DonZoomik]

I can't find the topic but it was mentioned a while back that you can manually move files and then rescan repository so Veeam would register the move.

[veremin]

Manual files' relocation followed by repository rescan should be sufficient in this case. Thanks.

[rogerdu]

Feature request!!

[veremin]

Migrate backup data from extent X, as soon as it gets filed up to certain threshold? If so, this feature is already on our radar. Thanks.

[mkretzer]

What about the backup metadata? All our repos have files from every backup job and so if we would just move the folder from one to the other scale out extend we would at least have a duplicated vbm file. So should we just leave the vbm, overwrite the vbm or delete both vbm (as we would do with corrupted metadata)

[foggy]

Metadata for the entire job chain is stored on every extent where there's a part of the chain, so you can leave it or overwrite it, it doesn't matter.