I am hoping someone can help with this problem.

 

I keep getting an issue where the dockers stop. If I try to start them I get "Execution error. Error 403". If I disable docker in settings then re-start it says "Docker Service Failed to start" I have in the past deleted the image file and re-created them all, but I have done this several times now, the last was 2 days ago. Sometimes it won't even create a new docker.img file and I have to wipe the cache drive.

 

The logs are full of BTRFS write errors on the cache drive. I can see no SMART errors on the drive and I ran the memory test for 20 hours early this week with no errors reported.

 

I've attached the diagnostics from this morning before I did anything.

 

Thanks!

 

tower-diagnostics-20190228-0708.zip

You have a 60GB docker image. If you have things correctly configured it is very unlikely you would ever need more than 20. You have corrupted your docker image and your cache drive.

 

There is a whole section in the Docker FAQ about "Docker Image Filling Up"

 

https://forums.unraid.net/topic/57181-real-docker-faq/

 

You must make sure that each of your applications are only writing to paths that are mapped. The most common mistakes are wrong upper/lower case (linux is case-sensitive) and using a relative path instead of an absolute path (must begin with /).

 

Set all of your dockers to not autostart, stop docker services and delete the docker image. We can reinstall your apps later using the Previous Apps feature on the Apps page.

 

It might be simplest to just reformat cache. You will probably need some further instruction for this but I will wait to see if someone has a better idea for fixing cache corruption.

 

There are checksum errors on the cache device, i.e., data corruption, I also see some segfaulting, you might want to run memtest for a few hours.

 

Also, if you're overclocking your RAM, don't, Ryzen is known to corrupt data with overclocked d RAM, and it's not always detected by memtest.

Thanks for looking at the logs and helping, I appreciate it.

 

I increased the docker image size a few weeks ago because I suspected it might be filling up. When I looked this morning, only about 15GB of it was being used, so I don't think it is that.

 

I had suspected it might be RAM related, because I added a 16GB stick 2 months ago, so I ran the memtest for 20 hours on Monday and it found no errors. The RAM is  not being overclocked but the new stick I recall has slightly different timings than the 2x 8GB sticks that I had initially installed. I am going to remove the most recent stick of RAM and see if this stops happening.