Hello guys

I've rebooted my server and as I wanted to start my array I got a disabled disk. How is that possible that it just came up after a reboot?

Unmountable - No file system (32) 
Added Smart reports I had from the disk and currently doing an extended test.

Thanks for any help

nas-smart-20170728-0115.zip

Edited July 27, 20178 yr by In0cenT

Disabled (red balled) and unmountable are 2 separate issues. Do you actually have both at the same time, or just unmountable?

Both:

 

Lots of CRC errors, probably there is (or there was ) a bad SATA cable, you should include the complete diagnostics.

Added the diagnostics and the new smart report.

I got the sata cable from my H220 seller.

nas-diagnostics-20170728-1804.zip

nas-smart-20170728-1805.zip

Problem occurred right at mount time, lsi driver is not very helpful regarding the type of error, I would power down, replace cables on that disk (or if you don't have spares trade cables with another disk just to rule them out), power back on, start the array and post new diags.

I just used a different cable from the other controller which was free and it still didnt mount. I then put the drive in my other drive cage (ICY BOX IB-545SSK) and it still didnt mount. Also I've repluged all cables on the controller side.

Logs when drive was in different cage: nas-diagnostics-20170728-2322.zip

Logs when drive was back in original cage with all cables checked if they fitted correctly: nas-diagnostics-20170728-2329.zip

 

Guess the drive died?
Thanks for your help!

 

Disk not mounting does not mean the drive failed - it means there is file system corruption and you need to run the appropriate recovery tool.   The SMART report for the drive suggests it is fine, but the red X indicates a write failed - probably because of a loose cable or something else external to the drive (and that is probably what caused file system corruption).

 

since the is drive currently ‘ref-balled’ then unRAID is not actually using the physical disk - instead it is emulating it using the combination of  other drives plus parity.   If you run the appropriate recovery tool it will almost certainly get the ‘emulated’ disk back to a good state and be able to mount the drive again.   You can run the recovery tools by putting the array into maintenance mode and then clicking on the disk name to get to the page for doing that.

 

Assuming the recovery tool works and the disk becomes mountable again then you are going to have to rebuild the physical disk to match the emulated one. 

 

Edited July 28, 20178 yr by itimpi

Just now, itimpi said:

Disk not mounting does not mean the drive failed - it means there is file system corruption and you need to run the appropriate recovery tool.   The SMART report for the drive suggests it is fine, but the red X indicates a write failed - probably because of a loose cable or something else external to the drive (and that is probably what caused file system corruption).

 

since the is drive currently ‘ref-balled’ then unRAID is not actually using the physical disk - instead it is emulating it using the combination of  other drives plus parity.   If you run the appropriate recovery tool it will almost certainly get the ‘emulated’ disk back to a good state and be able to mount the drive again.   You can run the recovery tools by putting the array into maintenance mode and then clicking on the disk name to get to the page for doing that.

 

Assuming the recovery tool works and the disk becomes mountable again then you are going to have to rebuild the physical disk to match the emulated one. 

 

Thanks for your response, would you mind to explain the recovery tools part in detail? There arent any tools installed by default are there?

You didn't mentioned your parity is invalid now, this complicates things. when did it happened? It was valid on the first diags.

I've reset the config once, probably was this?

OK, I think I see what you did, you should've mentioned this earlier, it's on the logs but I wasn't looking for it.

 

You did a new config and disk10 failed right away, parity is not valid, hence the unmountable disk.

 

Do another new config, assign all disks except parity, start the array, if any disk is unmountable grab and post new diags, if all disks mount, stop, re-assign parity and start the array to begin the parity sync.

 

 

No I rebooted my server without any issues. As it came back up disk 10 failed, I then tried to reset config and issue still persists.

Should I still go with your suggestion?

You should never do a new config with a disabled disk, try what I suggested, if it was a cable problem you'll be fine, it disk10 is really bad it's going to be complicated.

Oh, and disk10 was already disabled at boot time, meaning it was disabled by unRAID prior to that, it then got disabled again after the new config you did.

Edited July 28, 20178 yr by johnnie.black

Just now, johnnie.black said:

You should never do a new config with a disabled disk, try what I suggested, if it was a cable problem you'll be fine, it disk10 is really bad it's going to be complicated.

You are a bloody STAR! Disk is back up!

Do you mind explaining why unRAID still marked the disk as bad although I made a new config prior?

Thanks!

After the new config, there were disk read and write errors immediately after array start, hopefully from a bad cable, so the disk was immediately disabled, and don't forget that you never do a new config with a disabled disk, I believe there's a warning about just that on the new config page.