Today fix common problems has found that I am unable to write to my cache pool as well as my docker image. My docker image is on the cache pool so I assume that it is a problem with the cache rather than docker. 

 

My syslog has filled up after a restart inside around an hour with errors like this:

Sep 23 19:49:41 Hydra kernel: BTRFS error (device loop2): bdev /dev/loop2 errs: wr 238, rd 0, flush 1, corrupt 0, gen 0

 

I can't see any issues in the SMART status of the two cache drives and I recently recreated the cache after formatting the SSDs. Around a week ago I moved my SSDs off my SAS backplane and directly connected them to the motherboard, changing cables in the process.

 

Any help would be gratefully received. 

hydra-diagnostics-20190923-1921.zip

Cache files system is corrupt, best bet it's to reformat, if it keeps getting corrupt you might have an underlying hardware problem, also note that because of this bug cache pool is not redundant.

On 9/24/2019 at 7:30 AM, johnnie.black said:

Cache files system is corrupt, best bet it's to reformat, if it keeps getting corrupt you might have an underlying hardware problem, also note that because of this bug cache pool is not redundant.

I reformatted my cache pool and started again but 2 weeks later I have corrupt BTRFS leaves again. I have moved my cache drives from the SAS backplane to the SATA connectors on the motherboard, I have replaced the SSDs over the last year, the cables and the HBA (although this is now only used for the HDDs) is there anything more I can check on the system? I need the system to be stable.

Post current diags.

As best I can tell my flash drive just died so there may be some delay until I can get it back up and running.