Hello -

I have just 1 cache pool, with 1 cache drive in it. It is a 2TB NVMe drive. Today, I noticed that the main page is listing this as a 2.6TB capacity drive, with more than 2TB currently used. This may have something to do with an issue I experienced the other day, where this drive had become full without my knowledge, and it was preventing VMs and docker services from starting. I ended up having to manually transfer files off of this drive to another build, using terminal. I am not sure whether I bumped something during that process, or what. I also had recently updated to the latest Unraid.

Thanks,

Steve

haven-diagnostics-20250617-1538.zip

Edited June 17, 20251 yr by stev067

After more digging in the forums, I found this proposed solution. After performing the steps, the issue is not resolved, and I have attached new diags.

haven-diagnostics-20250617-1717.zip

Post the output from:

zfs list -t all

4 hours ago, JorgeB said:

Post the output from:

zfs list -t all

cachetwo 1.91T 443G 144K /mnt/cachetwo

cachetwo/Handbrake 1.28T 443G 1.28T /mnt/cachetwo/Handbrake

cachetwo/Unsorted 11.7G 443G 11.7G /mnt/cachetwo/Unsorted

cachetwo/appdata 5.78G 443G 5.78G /mnt/cachetwo/appdata

cachetwo/backups 5.78G 443G 5.78G /mnt/cachetwo/backups

cachetwo/domains 168G 443G 168G /mnt/cachetwo/domains

cachetwo/isos 5.07G 443G 5.07G /mnt/cachetwo/isos

cachetwo/media 10.4G 443G 10.4G /mnt/cachetwo/media

cachetwo/public 152K 443G 152K /mnt/cachetwo/public

cachetwo/system 24.4G 443G 24.4G /mnt/cachetwo/system

cachetwo/personal 416G 443G 416G /mnt/cachetwo/personal

3 hours ago, stev067 said:

cachetwo 1.91T 443G 144K /mnt/cachetwo

Stats are correct, 1.9TiB= 2.08TB, are using dedup or similar? Any cloned data will make zfs show more capacity then the actual device.

I'm not familiar with dedup, but I recognize seeing that word when I was investigating the drive (see attached image). It does seem that the 2.6TB reported capacity is coming from the sum of the allocations, which were both 1.30 earlier.

Maybe a clue, I seem to remember seeing a button that "upgraded" the drive to ZFS or something to that effect, while I was investigating another issue recently, and cilcked it. There were no immediate problems after that though, so I'm not sure if that would be related.

This is not a problem per se, like mentioned, if you have any deduped/cloned files on the pool, that is normal with zfs, if you don't know what is causing that, move data to the array or another pool, then move it back, if dedup is not enabled, the stats should then be normal.

1 hour ago, JorgeB said:

This is not a problem per se, like mentioned, if you have any deduped/cloned files on the pool, that is normal with zfs, if you don't know what is causing that, move data to the array or another pool, then move it back, if dedup is not enabled, the stats should then be normal.

Is the below output strange, though? According to this guidance I found, the column 'DEDUP' should not exist if DEDUP is disabled for this pool.

I really appreciate you looking at this. Whatever I did, if I caused this, I would just like to change it back to normal, where the capacity shows as 2TB.

Like mentioned, dedup or any other clones, for example, see a new zfs pool using a 120GB device, dedup is disabled, copied a single ISO to the pool:

Now after using the block clone feature to clone that ISO 20 times:

Note that the remain free space is the same, which is correct, since clones don't use any space, but I have now 20 x 4.84 GB files on the pool, so it shows 96.9GB as the used space, and since capacity equals used + free, you get the 207GB, once more, this is not a problem, and normal zfs behavior, with cloned data.

I am not intentionally using dedupe or any clones. How can I check and un-do that?

2 hours ago, JorgeB said:

if you don't know what is causing that, move data to the array or another pool, then move it back, if dedup is not enabled, the stats should then be normal.

Apologies, trying this now.

I am getting "Umountable: unsupported or no filesystem" when trying to re-start the array, after removing and re-adding the pool.

Edit: I found another forum post where it was suggested (by you) to re-format the drive at this point. I have done that, and am copying files back to the drive now. It defaulted to btrfs, and the capacity is back to normal. If docker and VM work fine after the file copy, all will be good.

Edited June 18, 20251 yr by stev067

And this seems to be resolved now. Thank you for your guidance in this thread and others.