Subject: Re: Log area on-disk for the journal
To: Michael van Elst <>
From: Bill Studenmund <>
List: tech-kern
Date: 10/24/2006 11:26:41
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Tue, Oct 24, 2006 at 07:07:38AM +0200, Michael van Elst wrote:
> On Mon, Oct 23, 2006 at 05:16:14PM -0700, Bill Studenmund wrote:
> > But that's the point of journaling.
> To me the whole point of journaling is to avoid costly fsck operations.
> You don't need to weaken file system consistency when you introduce
> journaling.

Well, that's not what it is to a lot of other folks.

Also, think hard about exactly what it would mean if we didn't do the=20
journaling as I describe. If we _don't_ remove all of the sequence points,=
we have both all of the sequence points from before and ALSO we have a=20
two-stage commit! So things are even slower!

On a journaled fs, the journal is an integral part of the fs consistency.=
I'm not coming up with a great way to express this, but the journal is as=
key a part of consistency as the super blocks or the inodes.

Take care,


Content-Type: application/pgp-signature
Content-Disposition: inline

Version: GnuPG v1.4.3 (NetBSD)