Replicas, Backup Copy and Object Storage

[CerebralStuntman]

Hi guys, I'm learning the ways of Veeam B&R but could do with some guidance on a backup regime I'm looking to implement.

Case #04284354

The primary remit is that require backups to be retained for 3 years. This must include;

Last 7 days
12 monthly backups
Once yearly backups.

DC1 - Contingency vmware environment
DC2 - Production vmware environment

I'm little short on storage in DC1 so I want to make use of the object storage feature to offload archived backups to S3.

Currently, I have a daily replica backup job that runs from DC2 that stores replica images on storage located in DC1. This job has a 7 day retention. 'Job A'. Fine no issue with that.

On a separate Veeam server located DC1, I have created a backup job - 'Job B'. This backs up the above-mentioned replica copies produced by 'Job A' onto a local backup repository. 7 days retention.

I now plan to create a 'Backup copy job' which is linked to 'Job B' which uses a scale-out repository as its storage destination. This job will contain weekly, 4, monthly 12, and yearly 3 backup archives. I would like these copies to be offloaded to S3 storage as quickly as possible. Not sure if there are recommended backup settings that may enable this.

Would someone be able to validate or rule out my plans and if this is possible perhaps point out if there is a better way of achieving the desired outcome or even point out some of the potential 'gotchas' I may be faced with.

I'm using Veeam B&R 9.5 u4.

Thanks.

[veremin]

Your plan looks good and seems to match perfectly your requirements. Thanks!

[oleg.feoktistov]

Hi Gary,

Looks neat. Just curious - are you intending to offload only GFS backup copies to s3 or all of your backup copies?

Thanks,
Oleg

[CerebralStuntman]

Thanks for taking the time to read my post

Oleg,

Regarding your question...

The primary intention is to offload the GFS backups to S3. Certainly, the 7 day replica backups will remain on local disk. The backup of the replicas are currently stored locally but I would ideally like to offload these to S3 as well. Not sure if thats easy to set-up and doesn't convolute the solution in any way. Assume I would need to set-up a separate 'scale-out repository' for that?

Thanks,
Gary

[oleg.feoktistov]

Assume I would need to set-up a separate 'scale-out repository' for that?

With your current VBR version - yes. I see no sense in keeping backup and backup copies for the same entities on one scale-out repository.

I guess, the main role of your backup copy job is GFS backups creation, right? And the goal now is to save space on storages in DC1?
VBR v10 could help you with that as there we introduced GFS logic for backup jobs and copy policy for SOBR's Capacity Tier.
Combining it all, you get:
- Backup job with GFS enabled and targeted to a SOBR.
- Copy policy enabled for SOBR mirroring backups to S3 on each run.

You can then mix copy/move modes and adjust operational restore window accordingly depending on your offload agenda.
One SOBR less, space is spared.

Cheers!