[SOLVED] 1 drive shows up unformatted, syslog attached

booting up the server today, disk12 shows up as unformatted.  The unraid server has been running fine for the pass two years with no problems and nothing has been changed on it.  I attached the syslog and am hoping someone can tell me to proceed.  Thanks a lot.

syslog.txt

UnRAID could not find a Reiser file system on Disk 12.  Is there any chance you have a syslog from the previous session?  Something bad must have happened then, perhaps a power spike?  It may have corrupted the starting sectors of the file system.

 

I would suggest trying to run a non-correcting parity check here, just to test the first 10% of the drive, but I don't think your UnRAID v4.4.2 had that feature.  I *think* there is a chance that if it was a power spike that corrupted the earliest part of the file system, then the virtual Disk 12 could be fine, even if the actual Disk 12 was corrupted.  In other words, the parity might not have been altered, when Disk 12 was altered.  The implication is that a rebuild of Disk 12 might be your best solution.  Otherwise...

 

Failing any other solution, you are going to have to use the Check Disk File systems tool.  You will start with the basic check, then it will almost certainly instruct you to use the more drastic versions of the reiserfsck command.

 

However, I think you should wait and see if others have any further thoughts and ideas.  I hesitate recommending upgrading at this point, with a 'missing' disk.  If you can access files on Disk 12 (the virtual one), I would take this time to save any important ones elsewhere.

I don't have any other syslog before the one that I posted.  I have it hooked up to a ups but the server was off which I think because the ups ran out of battery.

 

So to fix the unformatted drive 12 I should do a rebuild of the drive and if that doesn't work I should use the check disk file systems tool?

 

Would it be ok to pull drive 12 and hook it up to another computer to see if I can pull the files off the drive to save elsewhere?

Don't do anything yet, I'd like to hear other comments first.  Have you tried to look at the files and folders on Disk 12?  As you probably know, UnRAID simulates missing disks, and reconstructs the files and folders, so you should be able to access them and copy them off.  It's best though that you not write to Disk 12.

 

The UPS running out of battery power would qualify as a sudden power outage.  I suspect it happened while something was updating the Disk 12 file system.

Typically, if the file system is unable to be mounted, it wlll show as unformatted.

Before doing ANYTHING else, stop the array and make a copy of your "config" directory with it in the stopped state.  (It will let you return the the current state easily) 

 

If you then power down, unplug or unassign disk12, and then reboot you'll simulate disk12 by reading parity and all the other data disks.  If it still shows as unformatted you know the reconstructed file-system as reflected on the parity drive is also corrupted.  (and a reconstruction of disk12 will result in a reconstructed and similarly corrupted disk12 on the replacement disk)

 

If by some chance the emulated disk12 is fine, then you can stop the array once more, re-assign (or reconnect) disk 12 and let unRAID perform the reconstruction.

 

If it is not fine, stop the array, reconnect/reassign the drive, power up, and then proceed to check its file-system as described in the wiki by running

reiserfsck --check /dev/md12

 

Joe L.

Not sure if I did everything correctly, but I unassign disk12 then rebooted and it shows missing disk.  I then reassign disk12 , reboot and it still shows as unformatted.  My next option would be to run the reiserfsck correct?

But do you see files on Disk 12?  When you use a file explorer of any kind, does the file system on Disk 12 look correct?  The physical Disk 12 is not mounting because of corruption in the Reiser file system, but we are hoping that the virtual Disk 12 is still intact.  Please try to view the folders on Disk 12.

no i don't see disk12 at all.  I went to \\tower and see all the shares and other disk listed but don't see disk12

Not sure if I did everything correctly, but I unassign disk12 then rebooted and it shows missing disk.  I then reassign disk12 , reboot and it still shows as unformatted.  My next option would be to run the reiserfsck correct?

Try to start the array with disk12 physically disconnected, but still assigned, then see if you can see the contents of disk12. When it shows missing disk, it should still allow you to start the array and emulate the missing disk. If you already tried that, it wasn't clear in your post. It looked like you unassigned and reassigned without starting the array.

Not sure if I did everything correctly, but I unassign disk12 then rebooted and it shows missing disk.  I then reassign disk12 , reboot and it still shows as unformatted.  My next option would be to run the reiserfsck correct?

Try to start the array with disk12 physically disconnected, but still assigned, then see if you can see the contents of disk12. When it shows missing disk, it should still allow you to start the array and emulate the missing disk. If you already tried that, it wasn't clear in your post. It looked like you unassigned and reassigned without starting the array.

 

I not sure what was on disk12.  I save everything using the high-water method and the server would decide which drive to save files to.

Not sure if I did everything correctly, but I unassign disk12 then rebooted and it shows missing disk.  I then reassign disk12 , reboot and it still shows as unformatted.  My next option would be to run the reiserfsck correct?

Try to start the array with disk12 physically disconnected, but still assigned, then see if you can see the contents of disk12. When it shows missing disk, it should still allow you to start the array and emulate the missing disk. If you already tried that, it wasn't clear in your post. It looked like you unassigned and reassigned without starting the array.

 

I not sure what was on disk12.  I save everything using the high-water method and the server would decide which drive to save files to.

It would still mount the emulated drive if it is not corrupted, so see if it mounts, even if it has no contents.

Not sure if I did everything correctly, but I unassign disk12 then rebooted and it shows missing disk.  I then reassign disk12 , reboot and it still shows as unformatted.  My next option would be to run the reiserfsck correct?

Try to start the array with disk12 physically disconnected, but still assigned, then see if you can see the contents of disk12. When it shows missing disk, it should still allow you to start the array and emulate the missing disk. If you already tried that, it wasn't clear in your post. It looked like you unassigned and reassigned without starting the array.

 

I not sure what was on disk12.  I save everything using the high-water method and the server would decide which drive to save files to.

It would still mount the emulated drive if it is not corrupted, so see if it mounts, even if it has no contents.

 

Ok I disconnected the disk12 and booted up the server.  It shows missing drive and to check the check box and hit start.  I did that but disk12 still shows as unformatted and in windows explorer disk12 doesn't show up.

Not sure if I did everything correctly, but I unassign disk12 then rebooted and it shows missing disk.  I then reassign disk12 , reboot and it still shows as unformatted.  My next option would be to run the reiserfsck correct?

Try to start the array with disk12 physically disconnected, but still assigned, then see if you can see the contents of disk12. When it shows missing disk, it should still allow you to start the array and emulate the missing disk. If you already tried that, it wasn't clear in your post. It looked like you unassigned and reassigned without starting the array.

 

I not sure what was on disk12.  I save everything using the high-water method and the server would decide which drive to save files to.

It would still mount the emulated drive if it is not corrupted, so see if it mounts, even if it has no contents.

 

Ok I disconnected the disk12 and booted up the server.  It shows missing drive and to check the check box and hit start.  I did that but disk12 still shows as unformatted and in windows explorer disk12 doesn't show up.

You can then concentrate on fixing the file-system;  it is corrupted on both the physical disk and the emulated disk.

 

You can re-assign it.  It will re-construct it, but it will still be unable to be mounted, so will still show as unformatted in the web-management-display.

Not sure if I did everything correctly, but I unassign disk12 then rebooted and it shows missing disk.  I then reassign disk12 , reboot and it still shows as unformatted.  My next option would be to run the reiserfsck correct?

Try to start the array with disk12 physically disconnected, but still assigned, then see if you can see the contents of disk12. When it shows missing disk, it should still allow you to start the array and emulate the missing disk. If you already tried that, it wasn't clear in your post. It looked like you unassigned and reassigned without starting the array.

 

I not sure what was on disk12.  I save everything using the high-water method and the server would decide which drive to save files to.

It would still mount the emulated drive if it is not corrupted, so see if it mounts, even if it has no contents.

 

Ok I disconnected the disk12 and booted up the server.  It shows missing drive and to check the check box and hit start.  I did that but disk12 still shows as unformatted and in windows explorer disk12 doesn't show up.

You can then concentrate on fixing the file-system;  it is corrupted on both the physical disk and the emulated disk.

 

You can re-assign it.  It will re-construct it, but it will still be unable to be mounted, so will still show as unformatted in the web-management-display.

 

ok so i'm going to be using the reiserfsck command to fix it?

ok so i'm going to be using the reiserfsck command to fix it?

correct.

 

Since the disk is not mounted, no need to stop SAMBA or un-mount the disk (since it is not mounted)

 

Start by typing

reiserfsck --check /dev/md12

and let it guide you to fixing the corruption.

 

Post here if you have any questions at any time.

Ok after putting disk12 back in it completed it's thing and just got finished.  I ran reiserfsck --check /dev/md12 but it tells me that the reiserfs superblock cannot be found on /dev/md12.  Failed to open the filesystem.  It said that i should run this utility with --rebuild-sb.  So should I do this? Is this the correct command "reiserfsck --rebuild-sb /dev/md12"?

Yes, that would be the next step.  You need to let each reiserfsck guide you.

 

See here for the correct responses to its (--rebuild-sb ) prompts.  Remember to answer in exactly the same "case" as expected.

http://lime-technology.com/forum/index.php?topic=1483

 

Once it is done repeat the

reiserfsck --check /dev/md12

and see what it next suggests.

 

 

ran --rebuild-sb and completed.  Then ran --check and got this error "Bad root block 0. (--rebuild-tree did not complete)".

OK, now looks like you can run

reiserfsck --rebuild-tree /dev/md12

 

OK, now looks like you can run

reiserfsck --rebuild-tree /dev/md12

 

it ran the check and completed.  Everything looking ok now,  Thanks a lot for the help.  I guess this is a good time to update unraid.  From what I can figure out from looking around.  All I have to do is download the newer version and copy the files from the zip to the flash drive.

if the emulated drive was corrupt as well (and it looks like it was) I think you now have to do a parity check/correct to sync parity with the corrections to the new drive since it was unmounted and not in the array when you did the fixes.

 

someone please confirm.

if the emulated drive was corrupt as well (and it looks like it was) I think you now have to do a parity check/correct to sync parity with the corrections to the new drive since it was unmounted and not in the array when you did the fixes.

 

someone please confirm.

you are incorrect in this case.  Since the correction was done to the "/dev/mdX" device, parity was automatically updated.  In those rare cases whee the file-system repair is performed on the /dev/sdX1 device, then you are correct, parity must be re-calculated  (a parity sync made)

 

All reiserfsck corrections must be done with the file-system un-mounted.  (It will not let you make a repair if it is mounted)

 

Joe L.