Initially had what looked like a faulty sata cable, removed the disk and re-balanced back to the one disk.

Have replaced the cable and on adding the drive back in, it's all turned to custard.

Initial steps were:

Re-added drive to appdata cache pool, prompted for a format of the disk. Ticked Yes.

Array started and says the appdata cache pool is unmountable/readable fs. I thought it should automatically pick this up and set it up as raid1, it's done this for me in the past. Assume it as a v7 bug, which on googling looks like it might be?

Rebooted the array and then got this:

Wrong Pool State
appcache - too many wrong or missing devices

Found a few posts about this to remove the pool and re-add it back but in doing so when the array starts, it just comes back with no drive found in the 2nd slot(the original good disk with data) which is assigned to slot one but shows as missing from slot 2.

If i remove the pool again, I just get the same.

Edited March 10Mar 10 by Distorter

Attach Diagnostics ZIP to your NEXT post in this thread.

diag attached.

unraid-diagnostics-20260310-1558.zip

Even if I redo the Pool on reboot it has a mare:

19Z was in slot one whenever I create the pool and then on reboot it says it is wrong because it drops that disk completely on array start up.

The "wrong" is wrong or more so unraid is very confused.

Post the output from btrfs fi show

3 hours ago, JorgeB said:

Post the output from btrfs fi show

Label: none uuid: b87d8b7b-fbe0-4260-8dbd-05b70b3f9ce7

Total devices 2 FS bytes used 552.59GiB

devid 1 size 931.51GiB used 618.03GiB path /dev/nvme0n1p1

devid 2 size 931.51GiB used 618.03GiB path /dev/nvme1n1p1

Label: none uuid: a36906f0-74ea-4479-b57b-fee2e8aec608

Total devices 1 FS bytes used 440.00KiB

devid 1 size 1.00GiB used 126.38MiB path /dev/loop2

Label: none uuid: e4152190-aeeb-4c91-b1c1-6e222b394f50

Total devices 2 FS bytes used 77.58GiB

devid 1 size 465.76GiB used 84.03GiB path /dev/sdg1

devid 2 size 465.76GiB used 54.06GiB path /dev/sdh1

I've removed the pool so I can start the array and get to my shares if that changes anything in those results.

So I can successfully mount /dev/sdh1 from the unassigned device and browse this.

But mounting /dev/sdg1 give this error:

Mar 11 00:49:44 Ragnarok unassigned.devices: Mounting partition 'sdg1' at mountpoint '/mnt/disks/S2R9NX0J522893R'...

Mar 11 00:49:44 Ragnarok unassigned.devices: Mount cmd: /sbin/mount -t 'btrfs' -o rw,relatime,space_cache=v2,discard '/dev/sdg1' '/mnt/disks/S2R9NX0J522893R'

Mar 11 00:49:44 Ragnarok kernel: BTRFS info (device sdg1): first mount of filesystem e4152190-aeeb-4c91-b1c1-6e222b394f50

Mar 11 00:49:44 Ragnarok kernel: BTRFS info (device sdg1): using crc32c (crc32c-intel) checksum algorithm

Mar 11 00:49:44 Ragnarok kernel: BTRFS error (device sdg1): devid 2 uuid bac53136-eb52-4ea8-bf24-11b25eb40fe5 is missing

Mar 11 00:49:44 Ragnarok kernel: BTRFS error (device sdg1): failed to read chunk tree: -2

Mar 11 00:49:44 Ragnarok kernel: BTRFS error (device sdg1): open_ctree failed: -2

Mar 11 00:49:44 Ragnarok unassigned.devices: Mount of 'sdg1' failed: 'mount: /mnt/disks/S2R9NX0J522893R: mount() failed: No such file or directory. dmesg(1) may have more information after failed mount system call.'

44 minutes ago, Distorter said:

devid 1 size 465.76GiB used 84.03GiB path /dev/sdg1

devid 2 size 465.76GiB used 54.06GiB path /dev/sdh1

Pool appears to be out of sync. Try temporarily dropping the second device offline, is has less used space, so you are going to assume it's older:

echo 1 > /sys/block/sdh/device/delete

Then and without rebooting, create a new pool and add only sdg to it, start the array, and post new diags

OK, so that has at least allowed the fs to mount but I think the 54GB is correct as thats what it was originally before all of this. Not sure what the extra 30GB is? Or I could be wrong.

Things aren't working though as docker wont start.

unraid-diagnostics-20260311-0913.zip

Edited March 10Mar 10 by Distorter

Docker image is corrupt, delete and recreate:

https://docs.unraid.net/unraid-os/troubleshooting/common-issues/docker-troubleshooting/#re-creating-the-docker-image-file

Then:

https://docs.unraid.net/unraid-os/troubleshooting/common-issues/docker-troubleshooting/#reinstalling-docker-applications

Also see below if you have any custom docker networks:

https://docs.unraid.net/unraid-os/troubleshooting/common-issues/docker-troubleshooting/#restoring-docker-custom-networks

How would that break the pool and btrfs?

I tried to just go nuclear and did the new config and wiped the disks but it still won't mount the drives. This is as a brand new pool with wiped disks but it just loops and asks to format the drives over and over.

It's like unraid has blacklisted the disks?

unraid-diagnostics-20260311-2053.zip

Edited March 11Mar 11 by Distorter

3 hours ago, Distorter said:

How would that break the pool and btrfs?

Not sure what you are asking. The pool is currently working with one device, but the Docker image is corrupt and needs to be recreated. Once all is working, you can add the other device to recreate the mirror, but note that it will need to be wiped first, but only do that after confirming the pool is working as expected with the current one.

How do I get unraid to accept the disks as brand new like they've never seen them before?

If you mean pool SSDs you can wipe them with blkdiscard -f , all data will be deleted.

9 minutes ago, JorgeB said:

If you mean pool SSDs you can wipe them with blkdiscard -f , all data will be deleted.

Have tried that. Even did a new config.

Post new diags after trying that.

New diag attached. Re did commands above and then tried to format drives but unsuccessful.

unraid-diagnostics-20260313-1823.zip

Device is in use:

Mar 13 18:22:45 Ragnarok root: This disk is currently in use - repartitioning is probably a bad idea.

You cannot reformat a mounted device.

/dev/sdh1 466G 42G 423G 9% /temp_mount

Ah mounted it to pull the data. Will unmount and try again.

The gui just cycles over and over, doesnt give too much to go on.

Thanks for your help.

blkdiscard -f was the command needed to get the drive back into a working state.