Hi — I have a strange problem with the share appdata_kopia. This is the share where I keep backups of appdata and Nextcloud. The share settings are “primary cache, secondary array”, with “fill-up” enabled and it’s set to reside only on disk2.

A few days ago the problems started: overnight all folders on that share disappeared — only one remained — and the share owner changed from nobody to root.

I fixed it: I restored the backup and removed the container I thought was causing it (an Unraid config backup container — I don’t remember the exact name). Everything was fine for the next 2–3 days. This morning I started up and the share is completely gone. Strangely, yesterday at about 21:00 I manually ran a Duplicati backup from that very share.

Can someone please check the logs and tell me what’s going on? It worked fine for months. Alternatively, what can I check myself? Apart from Duplicati and an rsync job that runs in the morning, nothing has access or reason to touch that share — and deleting an entire share is a pretty major action.

The only thing I found in the logs is this fragment:

Sep 9 22:04:23 Arrakis shfs: /usr/sbin/zfs unmount 'cache/appdata_kopia' 2>&1

Sep 9 22:04:23 Arrakis shfs: /usr/sbin/zfs destroy 'cache/appdata_kopia' 2>&1

For some reason something removed that share — but only from the cache, which is quite odd.

arrakis-diagnostics-20250910-0537.zip

1 hour ago, Joloxx9 said:

Sep 9 22:04:23 Arrakis shfs: /usr/sbin/zfs unmount 'cache/appdata_kopia' 2>&1

Sep 9 22:04:23 Arrakis shfs: /usr/sbin/zfs destroy 'cache/appdata_kopia' 2>&1

This would be caused by the mover.

Just now, JorgeB said:

This would be caused by the mover.

I know, I have set mover to remove empty folders, but... Mover is scheduled to run at 4AM, and I didn't trigger it in the evening.

It might be worth running a check filesystem on disk2 in case file system level corruption is what is causing the share to not show up.

3 minutes ago, itimpi said:

It might be worth running a check filesystem on disk2 in case file system level corruption is what is causing the share to not show up.

OK running xfs_repair -n on them and see

Just now, Joloxx9 said:

OK running xfs_repair -n on them and see

Better to do it via the GUI as that ensures the correct device name is used. Click on the device on the Main tab to get to the check.

7 minutes ago, itimpi said:

Better to do it via the GUI as that ensures the correct device name is used. Click on the device on the Main tab to get to the check.

Done the test in maintenance mode, no issues found.

Edited September 10, 2025Sep 10 by Joloxx9

8 minutes ago, Joloxx9 said:

Done the test in maintenance mode, no issues found.

OK. If you look directly at the relevant drive(s) do files for the share show up?

1 minute ago, itimpi said:

OK. If you look directly at the relevant drive(s) do files for the share show up?

No, that's the thing.

Share was gone, files were gone.

Just to show you proper timeline, for months all was good, recently (last week) appdata_kopia share was wiped, I mean all files but one were gone. At the same time appdata_kopia share owner changed from nobody to root, I changed it back to nobody.

Since that time, for 3-4 days it was running okay until this night, when basically whole share with data vanished.

Unraid never deletes folders/shares on its own, so most likely some app/container did that.

I guess that is the case, but is there a way to find out somehow?

That usually won't leave anything logged, at least not on the Unraid syslog, if it was a container/plugin, it may on its own log.

Ok thanks, this might be extremely hard to track, but I think I might get an offender, nextclud borg backup.