Docker Service failed to start, Rebooting Temporarily Fixes it

[preventive-dungeon1911]

Hi All,

 

For some reason, Docker keeps crashing. After reviewing the system logs, the main thing that stands out is:

 

Oct 6 03:41:04 HS kernel: BTRFS error (device sde1): block=1077952512 write time tree block corruption detected Oct 6 03:41:04 HS kernel: BTRFS: error (device sde1) in btrfs_commit_transaction:2466: errno=-5 IO failure (Error while writing out transaction) Oct 6 03:41:04 HS kernel: BTRFS: error (device sde1: state EA) in cleanup_transaction:1964: errno=-5 IO failure Oct 6 03:41:05 HS kernel: loop: Write error at byte offset 992399360, length 4096.

 

I'm assuming the device sde1 is my main cache drive, since the path is /dev/sde in system devices, so could it be a faulty drive?

 

I attached my diagnostics. I'll keep digging around and post a solution if I find one.

 

Thanks!

 

**Edit**

I cleared my logs and stopped / restarted docker. For some reason the docker.img is in use and terminates the start.

Oct 6 08:53:32 HS emhttpd: shcmd (328): /usr/local/sbin/mount_image '/mnt/user/system/docker/docker.img' /var/lib/docker 100

Oct 6 08:53:32 HS root: '/mnt/user/system/docker/docker.img' is in-use, cannot mount

Oct 6 08:53:32 HS emhttpd: shcmd (328): exit status: 1

hs-diagnostics-20231006-0845.zip

Edited October 6, 2023 by preventive-dungeon1911

[JorgeB]

The pool filesystem went read-only, and this error:

 write time tree block corruption detected

can indicate a RAM issue, so good idea to run memtest.

 

There are also macvlan call traces, change the docker network to ipvlan.

[preventive-dungeon1911]

Set the network to ipvlan and memtest is running. Memtest is on its 3rd pass with 0 errors.

[preventive-dungeon1911]

So I ran memtest for a day or so and nothing failed. However, after changing the docker network to ipvlan and removing a potentially faulty array hard drive, everything is good again.