Hello friends,
I have a Thecus N8800Pro 2u 8 bay NAS. Some months ago I began troubleshooting an issue where the device becomes unresponsive when writing or reading files. A user alerted me that their transfer got stuck. I kill there session and attempt to rsync their files to a NFS mount of the NAS. A few files and it stops. After a while it times out.
The interesting part, when I look at the Thecus device. 3 of the 8 disks have solid activity light. I reboot the device, it works fine, we try accessing those files again via NFS and it slows to a crawl. Same three disks pegged.
It runs busybox so I am able to see a load of 20-50 but no CPU or memory contention. I ran e2fsck and found some errors. I unmounted and fixed them. Updated the firmware. The situation seemed to be resolved for a month but now it is back. Same three disks. dmesg listed a few "rpc-srv/tcp: nfsd: got error -32 when sending 140 bytes - shutting down socket" but I did not find anything useful form that.
Any other explanations for 3 of 8 disks to be pegged in a raid 6?
[link][11 comments]