Long Term Retention Policy Guidance

[NeedsMoreRGB]

Our environment is 95% VMware, 5% Physical. We are licenced for both Veeam Backup and VAO (Enterprise Plus). VMs live on an all flash HCI array, and backups are written to an all HDD NAS via Veeam / iSCSI. We currently have daily backups with a 14 day restore point policy, using Incremental and synthetic fulls. We are replicating (replicas) VM data to a second site.

-We want to implement a GFS like structure for the goal of long term retention. One option is backup copy jobs, but those seemingly have a higher disk cost due to the full backups (as opposed to synthetic). Another option in my mind was to stagger regular backup jobs, each with their own backup chain and retention policies. This would result in keeping fewer full backups, and is more reliant on incremental / synthetic fulls. I have an example of this architecture in Google Drive (see link below - click through tabs to progress forward in time). Is this compliant with best (or acceptable) practices?

https://drive.google.com/file/d/14OxTv9 ... t4SLl/view

Thanks!

[veremin]

The best practice dictates storing GFS restore points as full backups, so corruption of one restore point does not affect all all weekly/monthly/yearly restore points.

As to number of points to keep, this depends solely on your needs and requirements. Cannot really suggest anything for this matter.

Thanks!

[NeedsMoreRGB]

@veremin

Thank you for clarification! I've been scouring over the helpcenter.veeam and bp.veeam documentation, but have not found documentation related to this practice. Any chance you have a link? In my scenario, there are 3 separate backup chains, so corruption in any single chain would not affect the other two, correct?