Disk read error, disabled + parity error shenanigans

[Traxxus]

Came home to disk 2 disabled today. WD 6TB Red.

 

Disk is 7 months old, no problems (smart test attached, 6 months old though).

http://pastebin.com/armb6L3L Here is the log, the log goes back to the 2nd so I cut off some of it.

 

Additional problem is, last 2 parity checks came back with 1 error, from this month and from last month, a scheduled non-correcting parity check that runs on the first, I never ran another one to fix it because I've been so busy it just slipped my mind. I've now decided to make the monthly a correcting check, not sure if that's the best option overall, but for me it is apparently. 

 

A single parity error that was steady over a couple months itself is not a big deal I don't think, I'm inclined to think this disk read error is also not a big deal and just a glitch in the matrix, but both together are problematic. I do not have an extra disk, so as far I can tell, my only options are to force Unraid to trust the disk? A rebuild won't complete with the parity error present right?

 

I do recall this happening before, I'm 90% positive it was the same disk, though it was last year sometime so I can't be sure. I just rebuilt it on the same disk, ran tests and checks, and everything came out fine. 

smart_201411270723_WDC_WD60EFRX-68MYMN1_WD-WXC1H3436266.txt

[trurl]

Something not working with your attachment.

 

Go to Tools - Diagnostics and attach your complete diagnostics zip. This should always be what you give us when using V6.

[Traxxus]

Here ya go.

tower-diagnostics-20160511-0827.zip

smart_201411270723_WDC_WD60EFRX-68MYMN1_WD-WXC1H3436266.txt

[trurl]

This drive, WXC1H3436266, isn't responding. Maybe a connection issue.

[Traxxus]

Well I shut it down, reseated, restarted. A green notification popped up that said something along the lines of array has returned to a good state, something. Disk 2 is still reballed obviously though.

 

Since there is a pending parity error that was never fixed, a rebuild will 100% fail right? I'm thinking the only real choice is to force a new config and tell Unraid to trust it?

 

Included new diagnostic.

tower-diagnostics-20160513-0944.zip

[trurl]

Still not seeing any SMART for that disk. Instead it shows

Serial Number:    WD-WXC1H3436266
...
SMART support is: Available - device has SMART capability.
SMART support is: Disabled

SMART Disabled. Use option -s with argument 'on' to enable it.
(override with '-T permissive' option)

Not sure how to get it to report. Maybe someone else can say.

 

If parity isn't disabled but disk2 is disabled then normally you would rebuild disk2. And parity such as it is may still be better than the data on disk2 at this point. Do you know if anything may have been written to the emulated disk after it was disabled?

 

You should be able to look at the emulated disk2 data. That would show you exactly what would be rebuilt.

 

And I would like to see SMART for that disk if you intend to rebuild onto it instead of a new disk.

[Traxxus]

Last write was the morning before the error, it's a movies only share and it just hit the high water mark to move to the next, so it's out of rotation for a while.

 

I got smart enabled, used smartctl -a -A -s on /dev/sdh

 

Ran a short test on it, looks good I think.

 

Edit: Running a long test, I'll get back tomorrow. I do have a new drive ordered anyway, even if I don't use it for this always good to have a spare.

tower-smart-20160513-1308.zip

[Traxxus]

Long smart tests on all drives completed, attached.

 

 

tower-diagnostics-20160514-1132.zip

[Traxxus]

Well I got the new drive in, precleared, and rebuilt. It all went well, with the expected one error. I'll start a correcting parity check for overnight later.

 

The original drive is probably fine, I just wanted to do the rebuild on a separate drive so I would still have the data just in case something went wrong with the rebuild. I'll do a preclear or two on it then keep it as a spare.