Shared blocks on capacity tier

[desmith]

Looking into an unexpected increase in object storage utilisation for capacity tier in case 07011689

Retention policy and operational restore window/move to capacity tier configuration aims to offload only synthetic full restore points to capacity tier.

GFS retains 4 weekly, 12 monthly and 7 yearly backups, and capacity tier is moved to archive after 60 days.

I had thought the GFS restore points on object storage to share one copy of each unique block offloaded, but reviewing S3 folder size for a per-machine backup from a VM which I knew has had 2.5 TB of unchanged data between each restore point, and under 100GB of change between weekly backups at most, I see multiple folders over 1.6TB in size and about 1TB of data in smaller folders with modified dates in between the 1.6TB folders. Is this expected behaviour?

[haslund]

From version 12, indexes are no longer used during offload to capacity tier - so you are correct that this is expected behaviour now. You can see this documented at the top of this page: https://helpcenter.veeam.com/docs/backu ... ml?ver=120

[desmith]

Why the change from storage efficiency? From a quick Google search, AWS reliability figures for S3 are way above Audi's, but I remember there being a lot of 9s.

I'm good with keeping a few copies of each block, but actual fulls for every GFS in each tier?