The allocated space for my Zvol is significantly surpassed by its current usage.

[jhemley]

Hello, I'm running TrueNAS Scale as my SAN, with individual Zvols assigned to each SAN drive. One specific Zvol hosts my Plex VM, initially configured for 8TiB, yet it's consuming over 14TiB. I maintain backups in two locations, both on-site and off-site, making it quite costly. I'm puzzled about the excessive space usage, considering I've set it to retain snapshots for only 2 weeks. As far as I understand, snapshots should only occupy the space equivalent to the changes made to the drive. However, I haven't increased the Plex drive by 7TB in the past 2 weeks, so I'm unsure why this significant increase in size has occurred.

Additional note: The Plex VM is reporting its size in VMware as 6.5TB.

An update: I deleted all snapshots, which restored the capacity to 8TiB as expected. However, upon taking a new snapshot, the size instantly surged to 14.2TiB

 

[sretalla]

I deleted all snapshots, which restored the capacity to 8TiB as expected. However, upon taking a new snapshot, the size instantly surged to 14.2TiB

What's going on here is refreservation.

When you don't set a zvol to be sparse at creation, a refreservation size is set, meaning it will want the same size as the zvol to be reserved when taking a snapshot (to allow the entire zvol to change without filling the disk).

You can clear that with zfs set refreservation=none pool/zvol