Hello all,

Randomly yesterday My single cache drive (1tb nvme) became unmountable after some of my docker containers crashed and I restarted. The cache drive then kept coming up as unmountable, will btrfs errors in the logs. I restarted multiple times, and a singular time it worked again. I went to update my docker containers and it hung. I restarted and drive has been unmountable since. I pored over the previous posts related to this topic and looked on reddit to no avail. I have tried some of the command line fixes I have found for btrfs checks but nothing has worked. I apologize, i'm not the most knowledgeable about linux in general, and definitely not in command line. I setup this server 100% via youtube tutorials. I have attached my syslog and my diagnostics. Thank you in advance for any help you can offer!

homeserver-syslog-20251203-0047.zip homeserver-diagnostics-20251202-1944.zip

Type btrfs rescue zero-log /dev/nvme0n1p1 then restart the array and post new diags.

Wow! I ran that command and when I restarted it the drive was back online! I'm not sure if thats a permanent fix though, i've attached logs as requested and a smart report

Samsung_SSD_980_1TB_S64ANS0W512746B-20251202-0751.txt homeserver-diagnostics-20251203-0858.zip

The pool is detecting some corruption, scrub it and post the results from the GUI.

Also, the docker image is corrupt and needs to be recreatd:

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

awesome, thank you for the help. Right after I posted I realized that "Docker is unable to start" was still happening. Can you please elaborate what you mean by scrub it and post GUI results? I will need to backup my data first, as I do not have a redundant drive in the cache pool (lesson learned, and a 2nd drive has been ordered).

1 hour ago, prlahiff said:

Can you please elaborate what you mean by scrub it and post GUI results?

Click the pool, then "scrub"; when done, post the results.

Hello, I scrubbed, restarted, recreated my Docker image, and install plex as a test container. all appears to be well. Thank you so very much! I am still new to linux, do you have any insight as to what could have caused this? is it random bad luck, or potentially I have something misconfigured? I appreciate any resources or expertise you have on the matter. Again, you are the best TY!

homeserver-diagnostics-20251203-1849.zip

Interesting 2nd update. When I added back all of my main containers, they crashed and then it led to the unmountable issue again. I am going to reinstall clean docker image again and send a 2nd set of logs. I don't know if a particular container is the issue or if its something with the disk itself, or something else. Thanks again!

here is the 2nd set of logs. after restart, the drive is unmountable again. I did attempt to clean docker install and install some test apps, and containers did add but none of the apps worked correctly

homeserver-diagnostics-20251203-2022.zip

Since the log is corrupt again, I would recommend running the previous command again; it should make it mountable, then back up and reformat the pool.

Thank you. Looking through other posts I also found ram may be an issue. Lo and behold, 1 of my sticks is failing memtest with 1000+ errors, the other is testing no errors. I have ordered replacements. This ram is less than a year old is that common failure on unraid? also, could the ram have caused the cache pool corruption? I will backup and format drive, and also run without bad ram and see what occurs. Is there any official guide for the backup format and restore pool data to follow? Thanks for all the assistance

14 minutes ago, prlahiff said:

is that common failure on unraid?

Failing RAM has nothing to do with the OS, but I recommend using ECC for any important data.

14 minutes ago, prlahiff said:

Is there any official guide for the backup format and restore pool data to follow?

Not specifically for that purpose that I'm aware of, but you just need to copy the data, format, and copy it back.

OK thank you. I will do that after work and post a status update after some period of stability

it took a couple of days to get new ram that passed Memtest86. I re-ran the rescue command to make the drive show up. I then did the following:

A. Run the Initial Fix Command (Terminal in GUI)

You must run your specific repair command before the drive is accessible.

1. Click the Terminal icon in the Unraid Web GUI (usually the >_ icon at the top).

2. Type and execute your specific command to make the corrupted cache drive mountable.

3. Do not reboot after the command.

B. Disable Services

This is critical. Docker and VMs will lock files and prevent the Mover from moving everything.

1. Navigate to Settings $\rightarrow$ Docker. Set "Enable Docker" to No. Click Apply.

2. Navigate to Settings $\rightarrow$ VM Manager. Set "Enable VM Manager" to No. Click Apply.

C. Configure Shares and Run Mover

You must configure the shares that live on the cache (appdata, system, and potentially others like domains) to force them to move to the Array.

1. Go to the Shares tab.

2. For each share you want to move (e.g., appdata, system, domains):

   • Click on the share name to edit its settings.

   • Set "Primary Storage" to your Cache Pool name (e.g., cache).

   • Set "Secondary Storage" to Array.

   • Set "Mover Action" to Cache -> Array.

   • Click Apply.

3. Go to the Main tab.

4. Under Array Operation, click the Move Now button (or Move button).

5. Wait for the Mover process to complete. You can monitor the log for details. Once done, all data for these shares will be on your protected array disks.

When I try to do this, I am getting an error

Dec 9 18:09:18 HomeServer emhttpd: shcmd (131): /usr/local/sbin/mover start &> /dev/null &

Dec 9 18:10:43 HomeServer kernel: block nvme0n1: the capability attribute has been deprecated.

I am now unsure what to do next so I can properly backup the data on the cache drive to attempt a format to see if that can fix the btrfs corruption. logs attached. thank you again!

homeserver-diagnostics-20251209-1910.zip

7 hours ago, prlahiff said:

Dec 9 18:09:18 HomeServer emhttpd: shcmd (131): /usr/local/sbin/mover start &> /dev/null &

Dec 9 18:10:43 HomeServer kernel: block nvme0n1: the capability attribute has been deprecated.

Neither of these are errors, what is the current issue?

I wasn't able to get move to move stuff and see it on the array. I turned on mover logs and mover did in fact move them. I formatted the cache drive to ZFS, and repaired my docker image. when I reinstalled some of my apps, it now seems as that none of the data is remembered, but other apps had my settings preserved. For example plex does not seem functional, but gluetun does.. I think the original issue is solved. Would it be better to continue troubleshooting in this thread, or make a new one, as I no long have the unmountable or btrfs errors?

EDIT: it appears that container settings were preserved, but not settings with in the apps were preserved. I think this may be a fault of my own, as the settings in the apps were wiped when the cache was wiped. I understand this is not recoverable

Edited December 11, 2025Dec 11 by prlahiff

Do you have a backup of appdata?

I did not. lesson learned. I am reconfiguring my plex setup now. I thought appdata was backed up in the appdata folder, not the docker folders. I did think it was strange they disappeared when I formatted. Oh well. Its not like I lost any of my critical data that was on the array, Now i must reconfigure in app settings. lesson learned! unless you perhaps have another suggestion, but i looked through my shares and couldn't find anything that matched the data that would be needed for that

appdata is not backed up anywhere unless you back it up. There is a plugin for that.

Thank you both! I now have my containers running successfully again, all my appdata properly backed up and my Cache now on ZFS. Appreciate all the assistance to you both