Sure! Here’s a clean and detailed forum post you can use to ask for help on the TrueNAS Community Forums or Reddit (r/truenas):
Hi everyone,
I’m in a critical situation and would greatly appreciate any expert guidance.
What Happened
I have a TrueNAS SCALE setup.
My dataset is named /mnt/Data/Unica_Data and it is part of the Data pool (mirrored on 2 SSDs).
I created a Cloud Sync Task (Amazon S3) with:
Direction: PULL
Transfer Mode: SYNC
Path: /mnt/Data
Schedule: daily
Unfortunately, I made a mistake:
I misconfigured the S3 bucket (which had no files).
The sync ran with direction = PULL and sync mode, so it deleted everything in /mnt/Data to match the empty S3 bucket.
What I Tried So Far
-
I had automatic daily snapshots (Data@auto-2025-07-15_00-00, etc.).
-
I cloned the most recent snapshot:
zfs clone Data@auto-2025-07-15_00-00 Data/auto-2025-07-15_00-00-clone
- But when I checked:
ls -lah /mnt/Data/auto-2025-07-15_00-00-clone/Unica_Data
→ The directory exists but is empty.
- I confirmed the same across older snapshots — each one shows Unica_Data/ with no contents.
Additional Info
Snapshot size: ~22.8MB (Referenced)
Tried exporting the pool (zpool export Data) and re-importing readonly — no change
zfs list -t snapshot confirms consistent snapshot metadata, but no real file content is present in Unica_Data
Tried looking into zfs_recover, zfs-explorer, and zfspy, but:
Most tools are outdated or gone
Python tools fail with missing dependencies
Couldn’t find a working open-source deep ZFS recovery solution
Thank You
This is a critical production dataset, and I deeply appreciate any help or insights. If someone has been in a similar situation or knows a working ZFS recovery path, please guide me.
Thanks,
— Sohel