Unmountable disk present error

[storm123]

Hi all,

Bit of a panic. My array stopped and now is sort of back on line.

Biggest problem is that disk 13 is now reporting as unmounted and the main page of the web gui is giving me the option to format it. I don't want to do this as there's quite a bit of data on it.

I've done some searching for similar errors and pulled the diagnostic zip file (see attached). I've not done anything else.

In fact, I've just disabled the mover from kicking in shortly.

Any help on next steps would be appreciated.

Thanks,

tower-diagnostics-20161231-1844.zip

[JorgeB]

You need to check the filesystem on disk13 (md13)

 

https://lime-technology.com/wiki/index.php/Check_Disk_Filesystems#Drives_formatted_with_ReiserFS_using_unRAID_v5_or_later

[storm123]

Thank you Johnnie.

I've done as instructed and it doesn't look good.

I think I'll replace the drive and then try and work on the issues.

I may even have some warranty.

Should I just put a new drive in, pre-clear and then let it rebuild from parity?

[garycase]

Thank you Johnnie.

I've done as instructed and it doesn't look good.

I think I'll replace the drive and then try and work on the issues.

I may even have some warranty.

Should I just put a new drive in, pre-clear and then let it rebuild from parity?

 

Just to be clear, a drive doesn't "rebuild from parity" => it rebuilds using the data from ALL of the other disks to reconstruct the contents of the missing/failed drive.

 

You can try rebuilding the drive -- but if the drive has errors that are already reflected in parity, rebuilding will simply result in a mirrored copy of the bad drive.  If it's actually been disabled (red ball), then you can likely reconstruct it without a problem.  But it sounds like that's not the case.  It's easy to test:  Can you access the drive via Windows Explorer?  If so, then the system is emulating its contents;  if not, then the errors aren't going to be resolved by rebuilding the drive.    If you Stop the array; unassign the drive; and then Start the array, so it shows a "missing" drive #13, you will still be able to access its contents via Windows Explorer, as it will be emulated by the system.  This will show you exactly what the drive would look like if you did a rebuild.

 

[JorgeB]

It's not a disk problem (at least no apparently as SMART looks fine), you need to run reiserfsck with the --rebuild-tree option.

[AndrewT]

Thanks for your post. I just had a similar issue and resolved it by going into Maintenance Mode, then used:

reiserfsck --check /dev/md1

reiserfs_open: Your partition is not big enough to contain the

filesystem of (488381143) blocks as was specified in the found super block.

Failed to open the filesystem.

If the partition table has not been changed, and the partition is

valid  and  it really  contains  a reiserfs  partition,  then the

superblock  is corrupted and you need to run this utility with

--rebuild-sb.

reiserfsck --fix-fixable --rebuild-sb /dev/md1

reiserfsck --check /dev/md1

Before going off Maintenance mode and back to online.

 

 

[storm123]

Thank you for all the input guys.

Sorry for the delay in getting back.

Work and life has been so busy so just dealing with the issue this weekend.

I ran the fix-fixable option on that disk. I then ran the -check option again and found a couple of more errors. I then ran the rebuild-tree option. Again ran the -check and more errors. I was basically in a loop.

In the meantime, I was preclearing another 6TB disk. This kept failing in that the server would just lock up (no network/keyboard unresponsive etc.) and eventually I worked out one of the RAM sticks had errors. Using Memtest I identified the faulty RAM stick and it has now been removed and RAM wise it's all good now.

I'm just waiting for the rebuild-tree and preclear to complete.

What would be best next step?

Replace the drive or try starting the array once the check option confirms all is good?

Thanks again everyone.

[JorgeB]

When the rebuild tree finishes run another reiserfsck --check to confirm all is good now, I see no reason to replace the disk, SMART looks good, the problem was with the file system.