Hello, I've just run into a situation where we have some GFS jobs pointed to a repo that was not set up with per-vm backup files. This is non-standard for us but isn't causing any technical issue. The issue I'm running into is pulling the backup sizes per vm going to this repository. When I do this, it appears I'm getting usage numbers that more represent the entire job, not the single vm itself. Here is the code I use in normal situations that seems to work just fine:
Since a "storage" is the backup file itself, in this case containing all the vm backups, I guess it's expected. You should probably look at the task sessions for each vm to see what has been transferred/written for each restore point. Or switch over to per-vm if you have the space for new separate fulls.