Case #07608019: Backups using SOBR with performance tier extent sealed

[LEWISF]

Hello -

I hope I am posting this in the correct forum.

Our PRD site has SOBRs with a local NAS as the performance tier, and S3 buckets (AWS) as the cap tier. When we failover to our DR site, there is no local NAS but the site has the same access to the capacity tier. I need to know the impact of sealing the performance tier extent (NAS) and continuing to use the cap tier (AWS) for backups and restores.

My previous solution to this issue was to create buckets specifically for when we run out of our DR site and then reconfigure all the jobs to point to those buckets.

But since I can seal the performance tier, I need to know if I am correct in thinking that I can still run backups and restores to the EXISTING AWS buckets, or will that not work because Veeam will see those buckets as part of a SOBR and since there is no offload from the perf tier, the jobs would fail?

TIA.

Lewis

[veremin]

I'm not entirely sure I understand your scenario. Could you clarify it a bit? During the DR event, do you want to switch to a different backup server, add the existing repository used as a capacity tier to it, and start writing directly to this repository? Thanks!

[LEWISF]

Veremin -

I should have included that our Backup & Recovery server is replicated in Veeam to our DR location. So, during the DR event I would bring the replicated BRC server online. It will have access to the same SOBR resources that were in the primary site EXCEPT for the NAS. None of the SOBRs would have access to the NAS. There will be no NAS in the DR site so none of the SOBRs will have a performance tier. They will ONLY have a capacity tier. The capacity tier will be the same buckets that were in the PRD site.

I know I can seal extents. When we went through our first test run a few months ago for testing purposes I created separate buckets in the capacity tier specifically for when running out of the DR site. When I brought the replicated BRC server online in the DR site, I then reconfigured all of the jobs to point the storage repo for the job to the DR site repositories I created. Since I can seal a performance extent also, what I am not clear about is if I take an existing backup job, seal the performance extent in the SOBR it is using, can I still run backups / restores from the DR site? If that works, I would not have to reconfigure all 30+ jobs to point them to a dedicated bucket while in the DR site.

Let me know if this makes sense.

Thanks.

[veremin]

So, if I understand you correctly, you're asking whether on the replicated backup server, you can simply seal the non-existent performance extents and start running restores directly from the capacity tier, right? Thanks!

[LEWISF]

Correct. Restores and backups though. The Veeam engineer on my case said the backups would fail if I seal the performance tier so I would have to have separate buckets (cap tier only) for when we are in the DR location OR break the existing SOBRs to pull out the existing S3 repos that are in use. Then I would have to reconfigure all the jobs to point to the correct repo no matter which method I chose (dedicated buckets or the ones from the SOBRs). I don't think breaking out the S3 buckets from the existing SOBR will work though because there would be data in them that Veeam could not reconcile.

[veremin]

Restores will work smoothly. However, you won't be able to write backups directly to the capacity tier without using the performance tier.

Additionally, you cannot detach existing capacity extents and use them as the primary target for your jobs — they won't be able to recognize the backups already there and start writing increments to them.

Thanks!

[LEWISF]

Veremin -

Thanks. You confirmed what I was thinking.

Lewis