I'm pulling my hair out here and I'm hoping that someone can point me in the right direction. I routinely copy files to a folder on my Cache disk so that numerous Docker programs can interact with it. I generally have 70-80GB free on this disk however I'm running into problems where I copy 5-10GB over and receive an error on the copy that I'm out of disk space. When I check the disk space via the Unraid GUI I have the correct amount free, df command shows the same. After this my docker applications start failing as well and only a restart of the array fixes this issue until it pops up again.

 

Is there something that I'm missing? A setting perhaps?

If it's btrfs post the output of:

 

btrfs fi show /mnt/cache

 

btrfs fi df /mnt/cache

Thanks for the response, it is btrfs and here is the output:

 

btrfs fi show /mnt/cache
Label: none  uuid: xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
Total devices 1 FS bytes used 35.84GiB
devid    1 size 111.79GiB used 111.79GiB path /dev/sdg1

 

btrfs fi df /mnt/cache
Data, single: total=110.75GiB, used=35.27GiB
System, single: total=32.00MiB, used=16.00KiB
Metadata, single: total=1.01GiB, used=757.59MiB
GlobalReserve, single: total=256.00MiB, used=0.00B

This is a "characteristic" of btrfs, it allocates chunks before using them, and your device is completely allocated so it can't allocate a new metadata chunk, hence the out of space error, try doing what's in this thread:

 

https://lime-technology.com/forum/index.php?topic=56283.msg537129#msg537129

 

If deleting some files or -dusage=5 still gives and out of space error when doing the balance try deleting some more files or using a lower value.

Thanks johnnie.black. Unfortunately I don't have anything stored on there that is outside of my Docker appdata so I don't know how to clear anything without blowing away some of my Dockers.

Then your best course of action is to follow the replace cache procedure, but when it's empty format instead of replace and restore the data:

 

https://lime-technology.com/forum/index.php?topic=48508.msg516110#msg516110

 

Also make sure you update to latest unRAID, newer kernel alleviates this issue, but anytime used comes close to max size run a balance to prevent this from re-occurring.

 

devid    1 size 111.79GiB used 111.79GiB path /dev/sdg1

That's very helpful, thank you again for your quick replies!