Re: unusual superblock problem; crash debugfs

From: Eric Moors (scare.crow_at_oz.land)
Date: 12/23/04 

Date: Thu, 23 Dec 2004 07:56:59 +0100

clearthink@cavtel.net wrote:

> now what do I do?
>
> while creating backups my system paniced. i return
> to see blinking capslock, scrollock, numlock LEDs.
> i reboot.
>
> but cannot boot because my /dev/hda2 mount is unclean.
> i use my linux rescue CD. every attempt to run e2fsck
> on /dev/hda2 reports: "superblock corrupted" frags per
> block wrong.
>
> all attempts on backup superblocks reports: magic number
> wrong.
>
> an fdisk -l correctly reports all three partitions.
>
> so i run debugfs (version 1.32) on /dev/hda1. i run
> stats command. confirm block size is 1024.
>
> i re-run debugfs. i run command
>
> open /dev/hda2
>
> debugfs crashes with NULL pointer reference problem.
>
> anyway by assumping /dev/hda2 has same blocksize of 1024
> i retry e2fsck with every backup superblock known 8193,...
> and each one does not work.
>
> now what do I do? what can i do?

try with a proper backup superblocks and blocksizes:
If you have the sparse_super option set, the first backup
superblock is at 32768, the next at 98304 and so on.
Make sure to also specify the blocksize. Just try 1024, 2048 and 4096

eg. you may need: e2fsck -b32768 -B4096 /dev/hda2

Good luck.

Eric

Relevant Pages