On Wednesday 28 November 2001 12:32, you wrote:
> Declan Moriarty wrote:
> > I tried
> > mount /dev/hdb7 - bad superblock
> > e2fsck -p /dev/hdb7
> > It spits me back a notice saying "Try e2fsck -b 8193"
> > I tried for copies in 8K increments up to 49153, and 65537. Same result.
>
> Do you know what block size the fs uses? 8193 is only for 1k blocks.
> e2fsck(8) doesn't tell you where the alternate blocks are (except the first
> - 8193, 16384, 32768 for 1k, 2k, and 4k blocks) - it says to use mke2fs -n
> (the man page is here if you need it:
> http://linux.ctyme.com/man/man0437.htm). It makes sense that they are
> multiples of 8k, but you may be off by one here or there.
>
> You might try mke2fs -n before you go the -S route (and you need the block
> size for that too, anyway).
Well, this proved to be a smart move!! I hadn't the least clue what the
block size was, and could have gone to it's mirror image (/dev/hdb6) as a
last resort, but this did it even better.
mke2fs -n gave me a description of what the filesystem would be, INCLUDING
the adresses of the backup superblocks. The first one was on 98304, and that
let in e2fsck. Now I'm going to boot it, and see what's left. Mind you, the
main file entries destroyed were in the browser cache.
Thanks to both of you my butt is saved once again.
--
Regards,
Declan Moriarty
Applied Researches - Ireland's Foremost Electronic Hardware Genius
A Slightly Serious(TM) Company
Success covers a multitude of blunders - G.B. Shaw.
_______________________________________________
Linux-users mailing list
Archives, Digests, etc at http://linux.nf/mailman/listinfo/linux-users
