This has happened to my server 3-4 times in the past 2 weeks.  Issue with my cache drive(s)?  Getting a lot of BTRFS errors.  Currently have 2 running in RAID0. 

 

Haven't had to troubleshoot with the diagnostic logs yet so any help would be greatly appreciated!  Tired of rebuilding the docker.img

nickles-unraid-diagnostics-20201204-1207.zip

Edited December 4, 20205 yr by nicklesz

Looks like you filled up cache at some point and maybe that corrupted docker.img, but syslog is complaining about /dev/loop4, which isn't usual for docker.img, and in fact df is showing docker as /dev/loop3.

 

Delete and recreate docker.img then post new diagnostics.

 

 

Dec  4 06:58:39 nickles-UNRAID shfs: cache disk full

This immediately preceded the start of your issues with the docker image.  IIRC, BTRFS does not particular like (or handle well) drives becoming full, and very quick glance shows at least the media share writes to the cache drive until its 100%

You also need to run a balance on the pool since the filesystem is fully allocated, more info here:

https://forums.unraid.net/topic/62230-out-of-space-errors-on-cache-drive/?do=findComment&comment=610551

 

Thanks for the help guys! I rebuilt the docker.img and installed mover tuning to automatically move at 60% (forgot to do this after my USB drive failed recently).  Also did a balance on the cache pool.  Attached is the new diagnostics. 

nickles-unraid-diagnostics-20201206-1458.zip