Re: Ext2 defragmentation

Rogier Wolff (R.E.Wolff@BitWizard.nl)
Thu, 18 Nov 1999 10:20:25 +0100 (MET)

Theodore Y. Ts'o wrote:
> Date: Wed, 17 Nov 1999 07:45:32 -0500
> From: TenThumbs <tenthumbs@cybernex.net>
>
> > 8 meg block groups imply you're using 1k blocks. 32 meg block groups
> > imply you're using 4k blocks. E2fsprogs 1.18 automatically creates
> > filesystems with 4k blocks when making large filesystems, but most folks
> > have filesystems created with earlier versions of e2fsprogs which used
> > 1k blocks by default, and so there are many filesystems using 1k block
> > sizes out there. That's why I said 8 megs or 32 megs.....
> >
>
> Doesn't this change in e2fsprogs make it harder to recover data from
> a corrupt fs since there's no obvious way to know where to look for
> backup superblocks.
>
> Yes, there is that downside.

Huh? Isn't there a backup superblock always at block 131072? (Unless the
disk is smaller than 128M).

On the other hand, if the superblock is hosed, it's quite unlikely
that the "root dir" and other stuff near the beginning of the disk are
in good enough shape to recover the filesystem without "expert"
(human) help.

Oh, and let me say in public again:

1k blocks -> 8k blocks/group -> 8M groups
2k blocks -> 16k blocks/group -> 32M groups
4k blocks -> 32k blocks/group -> 128M groups

(That's why it REALLY helps in the "delete speed" and "fsck speed" to
have larger blocks.)

Roger.

-- 
** R.E.Wolff@BitWizard.nl ** http://www.BitWizard.nl/ ** +31-15-2137555 **
*-- BitWizard writes Linux device drivers for any device you may have! --*
 "I didn't say it was your fault. I said I was going to blame it on you."


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/