Just like the title says, occasionally, the docker service will "fail to start" even though it was already started and was working fine, then randomly, the service will fail crashing all applications.

The array will then proceed to go offline (even though its still running??? and says its its running mover/parity sync.) no data on the array is lost. This results in basic functions of the server environment to fail. Your unable to load parts of the GUI related to Docker, the system is unable to "stop the array" meaning that a normal reboot command (both in GUI and CMD fail to execute.) powerdown ;
Docker.img runs on a M.2 SSD ZFS Cache
i have uninstalled and reinstalled containers multiple times by this point.
i have attempted multiple repairs of the img files.

i have made new img files.

the only "fix" has been

echo 1 > /proc/sys/kernel/sysrq

echo b > /proc/sysrq-trigger

this kills the server like holding the power button for 10 seconds. thus rebooting from a clean slate.

My theory's,
1: Next cloud AIO corrupts the img file and results in failure of the docker service. (transitioning to a new self hosted provider now)
2: the Jellyfin container corrupts the .img file and results in the failure of the docker service (jellyfin is here becuase when the app backup plugin runs it can never verify the backup...)
3: RAM is bad (this was a windows machine prior to being a Unraid server. Ram showed no issues when running on windows 11. )
4: Server side bug with hardware configuration (maybe an update will help?)
5: Preserve user defined networks: Yes (is this the root cause?)

system specs.
Unraid 7.1.4
Core i5 9400f, Intel Arc A380 (for av1 decode.encode) GTX 1080TI (ollama), 32GB ram DDR4 ram running at 2133 (it can run 3200 but because of these issues i reduced it), 35 TB of drive space 3- 12'sTB 1- 10TB, 1- 1TB

Unraid installed on a 64Gb flash drive
Flash Vendor:

Verbatim

Flash Product:

STORE_N_GO 

The BTRFS Scrub with error correct checked.
UUID: dddddddddddddddddddddf

Scrub started: Thu Oct 30 09:18:49 2025

Status: finished Duration: 0:00:36

Total to scrub: 29.86GiB Rate: 849.27MiB/s

Error summary: csum=2 Corrected: 0 Uncorrectable: 2 Unverified: 0

Edited October 30, 2025Oct 30 by SP259

memtest is on the boot menu

Attach Diagnostics ZIP to your NEXT post in this thread.

nas-diagnostics-20251030-1016.zip
attached are Diagnostics.
This is a new docker image file, checking for corruption between each install. Currently there is no BTRFS corruption on the vdisk.
Ill run a Memtest and get back with findings.

Oct 30 09:16:51 MatthewNAS kernel: BTRFS info (device loop2): bdev /dev/loop2 errs: wr 0, rd 10, flush 0, corrupt 15, gen 0

Docker image has some corruption detected. After running memtest, whether errors are found or not, recreate the image:

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

Hi Yall, Memtest reported immediate errors, after testing each ram stick. 2 sticks of ram reported frquent errors. 1 was severely damaged, the other was not damaged as much but still reported errors on multiple runs. The other two sticks seem to be ok for now (need more testing time). For the time being I have pulled the bad sticks.
Hopefully this resolves the issue. Tomorrow I will Rebuild docker image file again. (thank god its soo EASY!)

Going forwards, I am assuming it is best to remain on BTRFS for the img file? I'm not sure what advantages or disadvantages Directory or XFS have?

TY Mods!!!!

Edited October 31, 2025Oct 31 by SP259
Spelling

6 hours ago, SP259 said:

I am assuming it is best to remain on BTRFS for the img file?

That would be my recommendation.