We have a Zpool called sata-mirror-10T. We have two drives (each 10T) in a ZFS Raid1 configuration as the dataset for Proxmox Backup Server. It's mostly full. I noticed that as I cleared out some VMs, the utilization percentage stayed fixed at 99%. This puzzled me until I noticed that the size of the Zpool itself also decreased. When I cleared out some VM snapshots, after a garbage collection of several hundred gigabytes, the size of the datastore apparently shrunk. At the start, there was 9.6TB usage. It's now at 8.6TB. PBS indicates that the total size is now just above 8.6TB, instead of 9.85 which is what we expected.
Does anyone have any insight as to what's happening here? I suspect it has to do with the subvolume /Backups.
Code:
root@PBS:~# zfs list
NAME USED AVAIL REFER MOUNTPOINT
sata-mirror-10T 8.92T 45.4G 96K none
sata-mirror-10T/Backups 8.92T 45.4G 7.83T /mnt/datastore/Backups
root@PBS:~# zpool list
NAME SIZE ALLOC FREE CKPOINT EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT
sata-mirror-10T 9.09T 8.92T 173G - - 80% 98% 1.00x ONLINE -