Hello all, I’ve recently run into this apparently well-known issue with btrfs where it has hit 100% filled and has changed itself to read only. I already took these steps:

https://forums.unraid.net/topic/61497-solved-cache-disk-full-shows-plenty-of-free-space/

and these... https://www.reddit.com/r/unRAID/comments/j99brd/docker_failing_with_readonly_file_system_error/

https://www.reddit.com/r/btrfs/comments/hwpcaj/full_filesystem_btrfs_read_only/

Unfortunately the (btrfs balance start -usage=75 /mnt/cache) 10 through 80 incrementing by 10 did not fix anything. Currently my dockers are still working except one. I turned a non-critical one off and tried turning it on and as expected just like in another forum that i cannot find now i get "error code 403". Then I tried deleting a docker and got "Image cannot be deleted, in use by other container(s)". Opened the terminal for unraid and gave the command "docker rm (docker container) which then told me that the filesystem is read only and how i found these articles to read.

These forums are saying stop all dockers, turn docker off and remove the docker.img to have it reinstall a fresh image.

https://forums.unraid.net/topic/103534-solved-docker-read-only-file-system/

https://forums.unraid.net/topic/54533-624-cache-drive-reporting-quotno-space-left-on-diskquot/#comment-532931

But they also say that I need to reformat my ssd. So before i stopped my critical containers like plex, I thought I'd ask for some direction as to what i really should do since there appears to be confusion in these posts. I’m wondering why btrfs is the default file system if it’s so fragile and as Jorge B says :"Yep, btrfs is kind of high maintenance, hopefully this specific issue should be fixed once we go to kernel 4.14." Since we are way past that kernal and this problem persists, if I need to reformat the ssd, 1) what’s the best way to do that and 2) how do I choose a better file system that’s not so high maintenance?

Please post the diagnostics.

My apologies.  Should have thought of that.  Here's the file and I hope I at least got that right.

bigbertha-diagnostics-20230312-1220.zip

Syslog stops on March 8, possibly due to mover spam, suggest disabling mover logging unless you are troubleshooting the mover, reboot and post new diags after array start, it would also be a good idea to update to v6.11.5 first, newer kernel is better at btrfs chunk managing.

So the event did actually occur on the 8th. I hesitate because I know I will be left with non running containers and even worse Downtime lol. I'm just trying to be as careful as I can since if I stop them, I have no functionality with the server. Is it possible that the diag file shows what could have occurred since nothing has changed since? 

Log shows the docker image going read only due to ENOEPC, but it ends a few hours later and there's no info about the underlying filesystem.