Hey guys,

All of a sudden, my server which has been running fine for 5 years decides to spit this error at me.

Both disks are in a shared pool in RAID0.
I've tried the advice in this thread here:
Cache: Unmountable: Unsupported or no file system (after update) - General Support - Unraid

However, when I go to restart the array after removing the drives from the pool, it says 'Wrong pool state: Cache - too many wrong or missing devices'.

I'm freaking out because I know I've got my Plex database stored on my cache and I don't want to lose it for all my remote users.

I've attached my diags.

What should I do? Are there any steps I should be taking now to recover/backup things ahead of any solution?

I'm not very good with Linux/Unraid, so please explain like I'm 5.

nas-diagnostics-20250523-2017.zip

Edited May 23, 20251 yr by Corvus

If the pool is raid0 you cannot mount it with a missing device, there are constant errors with cache1, replace its cables and post new diags and the output from btrfs fi show

7 minutes ago, JorgeB said:

If the pool is raid0 you cannot mount it with a missing device, there are constant errors with cache1, replace its cables and post new diags and the output from btrfs fi show

Sorry, what do you mean by 'output from btrfs fi show'?

Please explain like I'm a total newb. I have zero familiarity with linux or terminal or anything like that.

16 minutes ago, Corvus said:

what do you mean by 'output from btrfs fi show'?

Open a terminal window in Unraid and type that command, then post the results.

1 hour ago, JorgeB said:

Open a terminal window in Unraid and type that command, then post the results.

Ok, I replaced the cable (literally the hardest thing ever to do in that case) and it's now displaying all of my drives successfully with no error messages on main tab, and automatically started the array on startup.

Terminal says:

HOWEVER, when going to the Docker tab, it says 'No docker containers installed'(!!!)
Also in 'Fix Common Problems', it says this:

It doesn't look like it's full though?

I'm freaking out now. Did I just lose all my docker containers, and all associated settings/databases??

What can I do?

Diags attached.

nas-diagnostics-20250523-2017.zip

Edited May 23, 20251 yr by Corvus
Whoops, forgot to add terminal output

There's a new docker image, so that's why it's empty, the old one should still be there, and if it wasn't, you could easily recreate it, the main issue for now is that there are still errors with a pool device, and some stats are missing from the diags, post the output from btrfs fi usage -T /mnt/cache

2 minutes ago, JorgeB said:

There's a new docker image, so that's why it's empty, the old one should still be there, and if it wasn't, you could easily recreate it, the main issue for now is that there are still errors with a pool device, and some stats are missing from the diags, post the output from btrfs fi usage -T /mnt/cache

Here it is:

It is raid0, so if one device is failing, and it appears to be if the cables were swapped, you may lose some data, I would try to copy what you can to the array or another pool/device.

1 minute ago, JorgeB said:

It is raid0, so if one device is failing, and it appears to be if the cables were swapped, you may lose some data, I would try to copy what you can to the array or another pool/device.

Ok, so how do I do that if I don't have Krusader to use anymore?

I tried copying the 'appdata' directory on the cache pool to another folder on the pool which is shared on my LAN, but it said:

If the pool is read-only you can't change anything on it, have to copy to other storage (e.g. array, unassigned device...)

1 minute ago, Kilrah said:

If the pool is read-only you can't change anything on it, have to copy to other storage (e.g. array, unassigned device...)

I tried copying (as my previous post), but it gave me an error.
Plus JorgeB said there's still errors on it.
How do I check if it's read only?

Try copying to the array, not the pool.

Just now, JorgeB said:

Try copying to the array, not the pool.

Ok thanks, that seems to be working.

What's next after that finishes?

Remove the bad pool device and create a new pool with the remaining, or add a new one.

16 minutes ago, JorgeB said:

Remove the bad pool device and create a new pool with the remaining, or add a new one.

A new what?

Do you mean the drive itself is dead?
If so, how is this possible if it's letting me copy data from it?

Hey so it looks like the transfer progress is stuck.
It's been like this for over 2 hours now.

Is it finished? I'm worried to press cancel because I don't want to screw anything up.

Ok so I eventually cancelled it, but I figured out that it ended up copying everything in that folder over anyway.
The next folder was the 'system' folder, but as I went to copy that... It got stuck immediately. So I pressed cancel, and rebooted the server from the UI. As soon as it booted, now it can't find that cache disk at all! It doesn't even say 'unmountable', just 'not found'!

The system folder contains my docker.img file.
Am I now screwed?

You mentioned previously that I can recreate it. Can you please guide me through this? I'm going to do to the store today and buy two new cache drives. Sorry for all the questions, but I'm completely clueless here.

Edited May 23, 20251 yr by Corvus

Docker image can easily be recreated, no need to try and copy that, try to copy all the appdata you can.

6 hours ago, JorgeB said:

Docker image can easily be recreated, no need to try and copy that, try to copy all the appdata you can.

Ok well it looks like the appdata is on one of the array disks.
I've got two new cache drives (1tb each).
Now what?

Create a new pool, restore the appdata, recreate the docker image:

https://docs.unraid.net/unraid-os/manual/docker-management/#re-create-the-docker-image-file

Then:

https://docs.unraid.net/unraid-os/manual/docker-management/#re-installing-docker-applications

Also see below if you have any custom docker networks:

https://docs.unraid.net/unraid-os/manual/docker-management/#docker-custom-networks

1 hour ago, JorgeB said:

Create a new pool, restore the appdata, recreate the docker image:

https://docs.unraid.net/unraid-os/manual/docker-management/#re-create-the-docker-image-file

Then:

https://docs.unraid.net/unraid-os/manual/docker-management/#re-installing-docker-applications

Also see below if you have any custom docker networks:

https://docs.unraid.net/unraid-os/manual/docker-management/#docker-custom-networks

Thanks.

When you say 'restore the appdata', how do you mean exactly? It's been around 4 years since I set this up. All I can remember is that the appdata lives on the cache and doesn't get 'backed up' to the array. Assuming I've got the appdata folder on disk 6, what's the best way to do this? Sorry for the silly question.

You just need to move it to the new pool, you can use the mover:

https://docs.unraid.net/unraid-os/manual/shares/user-shares/#moving-files-from-the-array-to-a-pool-cache

Ok, so should I replace the two cache drives with the new ones first, before I make a new pool?

Also, should I format them to ZFS? I'd like to maximize performance as much as possible, so would RAID0 still be the best choice? Or would combining them into a pool be just as good?
Do I need to do anything in the Unraid UI before I physically replace them?

Edited May 24, 20251 yr by Corvus

ZFS is a good option, you can use a stripe for best performance, but if one SSD fails you can lose the whole pool, so should have good backups.

40 minutes ago, JorgeB said:

ZFS is a good option, you can use a stripe for best performance, but if one SSD fails you can lose the whole pool, so should have good backups.

Ok so I mounted both drives in a single cache pool but mirrored. I think I learned my lesson this time. The cache pool is now correctly reading 1TB.

However, now there's another problem. During this process, I received a notification in the Unraid UI that disk 6 is unreadable.

So I shut down, reseated connectors, and rebooted.

Now theres a red X next to where the drive should be in the array, and hovering over it says 'device is disabled, contents emulated'.

I also notice that the icon for the Unraid chrome tab is an orange triangle instead of a green circle.

The drive now appears in 'historical unassigned devices'. I've swapped SATA/power cables to no avail.

Along with the two replacement cache SSDs, I also bought a 12TB HDD with the intention that I would eventually replace the 8TB parity drive, and the 8TB will replace the 3TB drive in my array. If in the event that disk 6 (which is a 6TB drive) is dead, I guess I can replace it with the 8TB ex-parity drive. But is this actually the case? If so, I'm in disbelief that two of my drives (one an SSD, and the other an HDD) could die within 24 hours of each other!

The array now takes forever to start upon boot.
How do I troubleshoot this? I really hope it's not the SATA port on my motherboard...

Diags attached.

nas-diagnostics-20250524-2239.zip