System recently finished a parity check (fixing 4305 errors) and my new chassis just arrived.

Clicked Stop Array in the WebGUI and now it is just sitting for awhile on "Sync Filesystems..."

How long should this process take?

I really do not want to go through another 15 hour parity check if it does not stop cleanly.

SSH in and top shows:

rsyslogd running around 153-156% CPU load

umount running at 100.0% CPU load

Also, Docker tab does not show up when I click on it during this Sync Filesystems nor does the Logs pop-up show anything.

tower-diagnostics-20170817-1059.zip

Update: "Sync Filesystems..." went away after 45 minutes but array still shows as started.

Clicking Stop Array now just takes me back to Main page refresh after about 2 minutes. Array still shows Started.

Going into Settings -> Docker says "One or more paths do not exist."

Does this mean the array IS shutdown since it cannot see the cache drive? (/mnt/user/system/docker/docker.img) or (/mnt/user/appdata)

Maybe some halfway state...

Edited August 17, 20178 yr by blurb2m
Additional info.

Make sure nothing is accessing the array, including a telnet session on a mounted folder, if nothing is open type reboot on the console.

@johnnie.black I had tried a shutdown and that killed all access to the server. (Had to hard boot) No telnet session was running.

Just got everything moved over to the new chassis (Norco 4220) and booted up. 

Array reads "STOPPED. UNCLEAN SHUTDOWN DETECTED."

What should my first step be? Start the Array and let it parity check or should I do a diskcheck on the cache disk?

Start the parity check, previous log is completely filled with call traces, if any more issues grab new diags.

Cache disk is btrfs. Can I run a cache scrub before the parity check? Trying to put off that parity check since it takes so long and it had just finished a 19 hour check before I tried shutting down...

UPDATE: Parity started itself. Can I scrub during parity check?

I read somewhere that someone was having similar issues and a scrub helped...

Edited August 17, 20178 yr by blurb2m

You can scrub at the same time, you can also cancel the check and start it later, but it should be done as soon as possible as it's normal to have sync errors after an unclean shutdown.

Just follow these steps for BTRFS?

https://wiki.lime-technology.com/Check_Disk_Filesystems

"Checking and fixing drives in the webGui" section

Click on cache and there's a scrub button

Rgr. Scrub completed with no errors. Will update tomorrow once parity finishes.