Hey there everyone. I seem to be having some trouble with my cache drive(s) on my Unraid server. All of my dockers suddenly complained of a read-only filesystem and I'm receiving BTRFS error messages. I'd provide the diagnostics file, but it won't let me download it (gets stuck at "please wait..."). Apparently I needed to be in maintenance mode for that to work. I can go in and grab specific files for anyone who wants to help out. Just let me know what information you need. I am attaching the errors and warnings portion of the syslog and a BTRFS filesystem check.

 

syslog warnings and errors.txt

BTRFS Check Status Maintenance Mode.txt

 

I'm ready to work the problem, just need some steps to follow so I don't mess anything up. Thank you to those who lend me their time, I really appreciate it.

 

Dell r710 Sever Chassis

I have a 120GB SanDisk SSD and a Hynix 512GB NVMe SSD (connected via PCIe) totaling 632GB.

 

OUTPUT OF "btrfs filesystem df:"

Data, RAID1: total=11.00GiB, used=555.85MiB

Data, single: total=88.00GiB, used=42.33GiB

System, single: total=32.00MiB, used=16.00KiB

Metadata, single: total=2.00GiB, used=269.98MiB

GlobalReserve, single: total=77.45MiB, used=0.00B

 

Getting this error often:

Warning [TOWER] - Cache pool BTRFS too many profiles (You can ignore this warning when a cache pool balance operation is in progress)
PC300_NVMe_SK_hynix_512GB_FS6BN1188103ABG0O (nvme0n1)

 

tower-diagnostics-20191024-1051.zip

Edited October 25, 20196 yr by hypercoffeedude
Added more information

Anybody?

@johnnie.black is the BTRFS guru, but offhand I'd say your first step would be to copy everything that's currently on the cache to an array drive or make sure you have it backed up somewhere. You may end up having to reformat the cache pool and put everything back.

Missed original post due to long time forum bug, but basically you need to do what jonathanm mentioned, filesystem is corrupt and with btrfs best way forward is to backup, reformat and restore the data.

 

 

Thank you, I did end up re-creating the cache pool. I figured I'd probably have to wipe and make fresh, but I was just curious if there was another easier way to solve it. Thanks @jonathanm and @johnnie.black