1

I have a Synology NAS using SHR mode, which is basically LVM and Raid 5. Somehow my LVM volume became corrupted and won't mount anymore. This volume houses over 20TB of data, and most of it is NOT backed up -- a lot of irreplaceable data like baby pics and videos, etc.

When I try to mount the volume manually after doing a fsck.ext4, I get the following error in /var/log/messages:

Aug 11 19:01:28 Uranus kernel: [ 816.184841] EXT4-fs (dm-1): ext4_check_descriptors: Block bitmap for group 163584 not in group (block 18343897062229723040)!

Aug 11 19:01:28 Uranus kernel: [ 816.196072] EXT4-fs (dm-1): group descriptors corrupted!

I just don't know what to do at this point. I think I would have more options available to me if I moved over the 12 drives to a Linux box but that is just not doable for me right now for obvious reasons.

1
  • Before doing anything with a drive, I would make a low-level copy of that drive (using dd). Then you can freely experiment with the copy, without risking losing more data. Feb 13 '15 at 14:22
0

First, get a list of the backup super blocks:

dumpe2fs /dev/YOUR_DEVICE | grep Backup

Then take one super block you just pick and run another fsck:

fsck.ext4 -b YOUR_BACKUP_BLOCK_NUMBER /dev/YOUR_DEVICE

This might do the trick. If not, time for the next step, but only if you are feeling brave:

mkfs.ext4 -S /dev/YOUR_DEVICE

This will rewrite all super blocks and group descriptors only. If you use it wrong changes are high you loose all of your file system contents., so read the man page before!

Oh an another word of advise: a RAID is no backup. Implement some kind of backup strategy after that!

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy