Offload to SORB using Incremental Transform

[RogerRabbit]

Hi,
I read a lot, but i´m a bit confused.

I currently use the forever incremental backups with Transform previous backup chains into rollbacks.
I do the transform every day (not once a week). It´s to release the snapshot faster, and I waste space with 2 VBK (If I use config the job without transform).

CT is Wasabi.

Job 1 (Platinum)
Retention policy: 30 days
GFS: 4 Weeks / 12 Months / 5 Year
Copy policy Disabled on SOBR CT.
Move policy enabled on SOBR CT with 30 days window.

In job 1 i understand that all restore points after 30 will be offload to Capacity Tier. In this case, all GFS and will be all restore point will be "FULL". It´s correct?

Job 2 (Gold)
Retention policy: 45 days
GFS: 3 Months
Copy policy Disabled on SOBR CT.
Move policy enabled on SOBR CT with 30 days window.

I´m concerned about what HannesK say in object-storage-f52/veeam-and-wasabi-t64162.html#p384567, I´m not using Reverse Incremental, but Transform.

I want restore from Capacity Tier indenpendent, if the performace Tier is lost. I know if i use the "Copy policy Disabled on SOBR" i will copy all restore point in CT, but i can i do it without Copy enabled?

[Gostev]

1. No, that is not correct: object storage offload is forever-incremental regardless of how you store your backups on-prem. No restore points in object storage will ever be independent full, including GFS backups. Any restore point is effectively a virtual synthetic full pointing to a set of objects from a forever-incremental chain. There's no difference whether on-prem, a restore point was stored in an incremental backup file or in a full backup file. All restore points are the same in object storage.

Keep in mind that all the complexities of backup modes and backup chains with on-prem storage are solely due to the fact that we have to encapsulate restore points into files, before we can save them onto a file system. Object storage does not have a file system and so such problem, so we don't have to deal with all these complexities there. Any restore point is just a little metadata that describes what objects it consists of.

2. I don't understand the question. You make a confusing statement there: with the "Copy policy DISABLED on SOBR", of course it will be the opposite of what you said - there will be NO "copy all restore point in CT".

If you're asking whether you can restore from the Capacity Tier with only the Move policy enabled, then the answer is yes. However, keep in mind the latest backup available for restore in the Capacity Tier will be 30 days old in this case (based on your offload schedule). So, if the Performance Tier is lost, you will lose 30 days worth of your most recent backups, and will only be able to restore very old data.

The only way to protect yourself against the loss of Performance Tier is to have the Copy policy enabled.

[RogerRabbit]

Thank you for answering.

I got confused in my question but you answer clearly.

So I will use:
- Incremental backups with Transform previous backup chains into rollbacks. (Release the snapshot faster)
- Copy policy ENABLED on SOBR. (Allow restore, if I lost performance Tier)
- Move policy enabled on SOBR CT with 30 days window. (It will delete from Performace Tier, all restores point old than 30 days)

It´s a good config?? Any suggestions?

[Gostev]

Yes, all good - except I recommend against using the "Transform previous backup chains into rollbacks" option:
1. It does not release the snapshot faster comparing to regular incremental backup. In both cases, the transform process only starts after the snapshot has been already released.
2. This option is deprecated in v11 and will likely be removed completely in v12.