kjclark23

I had to install the production branch driver, which is 580 vs 590. Doesn't look like my card will be supported in future.

Just to update, I edit my Emby docker and got rid of the extra parameters "--runtime=Nvidia" and the docker runs fine. I got to put the extra parameter back and it fails again. Now I'm trying to figure out what causes this to happen. Hoping it's not having to wait for another driver update.

I deleted the vDisk and created a new one. Went to add the Dockers and most ran successfully, but some didn't. I noticed that it has something to do with Nvidia. I copied some of log for the Emby docker below. But each docker that uses the video card fails their docker build. docker: Error response from daemon: failed to create task for container: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error running prestart hook #0: exit status 1, stdout: , stderr: Auto-detected mode as 'legacy' nvidia-container-cli: initialization error: nvml error: driver not loaded: unknown. The command failed. So I'm guessing this is an Nvidia driver issue now. I checked and I have the lastest driver. gamingnas-diagnostics-20251222-1403.zip

I updated from 7.2.2 to 7.2.3 today. First, the update wasn't smooth, as my server froze when rebooting. I was able to see that it was stuck in the loading screen where you can enter the bios. After shutting off and turning back on, I was able to see Unraid going through the rest of the update. The server booted back up, but now I have Dockers not starting, giving "Execution Error". Wanting to post this in case there's something going on with the update. It of course could be my issue. I also have no clue what to do here. I did a scrub, with no errors (not sure if this even helps). I had a docker I didn't regularly use do this before, I just waited and it corrected itself. However, I don't want to wait for these Dockers since they're Emby and Plex, which gets used regularly. Any next steps would be appreciated. Posted diagnostics. gamingnas-diagnostics-20251222-1233.zip

Thank you! Back up again. I will backup and reformat it.

All of a sudden my server crashed. I had to manually restart it and now my cahce drive is back to unmountable again. Ugh. I tried doing what you said before with "btrfs rescure..." but that resulted in "No valid Btrfs found on /dev/sdr1 ERROR: could not open ctree" Posting diagnostics. Starting to worry I have a hardware problem. gamingnas-diagnostics-20251005-2115.zip

Many thanks. The scrub came back with no errors. Now I just need to restore my previous containers. Thanks for holding my hand!

I disabled the docker and deleted the newer image file. I ran the scrub and got this output UUID: 6738fddc-e7e7-4acf-bf7a-c650d6ca514a Scrub started: Thu Oct 2 13:12:27 2025 Status: finished Duration: 0:07:09 Total to scrub: 455.63GiB Rate: 1.03GiB/s Error summary: csum=8 Corrected: 0 Uncorrectable: 8 Unverified: 0I then went to enable the docker again and I get Docker Service failed to start. Attached more diagnostics. gamingnas-diagnostics-20251002-1325.zip

I'm seeing I have two docker image files, caused by Unraid creating a new one when not finding the old one. I assume I should just delete the newly created one to fix this? Also, should I disable or enable the docker before I do this?

Awesome, that got them recognized now. And my appdata is back since they lived on my cache drive. However, my docker containers are still gone, not sure what happened to them. I believe I should be able to recover those, forgetting how. Attached diagnostics after starting the array. gamingnas-diagnostics-20251002-1228.zip

I just want to make sure I understand the steps. Assign the pool devices back Type "btrfs rescue zero-log /dev/sdr1" in the terminal Start array Is this right?

All my docker containers suddenly disappeared and noticed my cache drive was Unmountable: Unsupported or no file system. I have tried stopping the array and removing the cache drives (pool of 2) to start the array but then I get "Wrong Pool State Cache too many wrong or missing devices". I have attached diagnostics. I'm hitting a brick wall. gamingnas-diagnostics-20251002-1102.zip

I went and changed the /storage back to "ro" and MakeMKV is back to stating it is in read only mode and I get an error. I'm finally seeing, thanks to your post, that MakeMKV is somehow defaulting to using the /storage for its output. Not sure how this automatically happened, but I changed it to /output which has "rw". Seems to be working now, really sorry to waste your time on this. I will always be learning with this stuff.

Do I understand docker volume mapping? I can't really explain it myself, so no. I didn't think I would want the /storage to "rw", but changed it to see if the docker would then work. For the path, I have it mapped within MakeMKV GUI to send to just /storage (without /user). I guess I needed to change this in the settings too. I didn't originally have it this way, use to work for a long time. MakeMKV gave me a pop-up stating that the path doesn't exist, which I thought was odd. So I changed it in the GUI to get it working. Not sure if this has anything to do with updating to Unraid 7.

I see this now. This is interesting. Can these settings get changed on their own? I did not change this and wonder why it didn't work, lol. Curious how it has been working for such a long time. But I changed the /storage to "rw" and fixed the /mnt to /mnt/user. I'm really not sure if the /mnt was always like that or if I had the default (which I usually keep). Just really not sure how this even happened.