BTRFS and Loop2 Errors

[MagicMountainMan]

Earlier today my BTRFS cache drives went into read only mode as the log spit out a bunch of BTRFS errors regarding loop2. From what I've gathered searching around is that my Docker image is most likely corrupted and that it could be my RAM that's failing. I'm currently running Memtest86 to see if it's indeed the RAM, but I thought I would post my diagnostics here so that someone much smarter than me could have a look through and let me know if it could be something else.

 

Thanks for the help!

tower-diagnostics-20220906-0805.zip

[JorgeB]

It does look like the typical bad RAM scenario, but you appear to be using ECC RAM, so technically it should not be possible, could be some other kernel memory corruption.

[MagicMountainMan]

9 minutes ago, JorgeB said:

It does look like the typical bad RAM scenario, but you appear to be using ECC RAM, so technically it should not be possible, could be some other kernel memory corruption.

I am indeed using ECC RAM. So far Memtest has been running for a while and nothing has failed. Not really sure where to go from here.

Edited September 6, 2022 by MagicMountainMan

[trurl]

45 minutes ago, MagicMountainMan said:

Docker image is most likely corrupted

Are you sure you didn't fill it or the drive it is on? Why do you think you need 40G docker.img?

[MagicMountainMan]

8 minutes ago, trurl said:

Are you sure you didn't fill it or the drive it is on? Why do you think you need 40G docker.img?

Yeah the drive was only a little more than half full when I logged in. 40GB Docker image I think was from a long time ago when a container had runaway logs and it kept filling up the image. I just don't think I ever changed it back.

[MagicMountainMan]

I ran Memtest for 8 hours sans haven't gotten a single error. Booting back into unRAID now to see what else I can do. Not really sure what to do at this point.

[JorgeB]

Run a correcting scrub on the pool and post the results.

[MagicMountainMan]

4 uncorrectable errors, this is the second scrub I've done though. Before this I also did a filesystem check and it found nothing...

[JorgeB]

Please post new diags if you didn't reboot after the scrub.

[MagicMountainMan]

Diagnostics attached.

tower-diagnostics-20220907-0915.zip

[JorgeB]

(path: appdata/plex/Library/Application Support/Plex Media Server/Media/localhost/b/f670656af31cc2920ab8d8918bd81f105ae352a.bundle/Contents/Chapters/chapter3.jpg)

 

This is currently the only corrupt file, delete/restore from backups and the pool should be fine for now, but if you get more corruption there's likely an underlying hardware issue, you can also take a look here for better pool monitoring.

.