Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 07 Jun 2005 22:25:58 +0100
From:      David Malone <dwmalone@maths.tcd.ie>
To:        Scott Long <scottl@samsco.org>
Cc:        scottl@FreeBSD.org, hackers@FreeBSD.org, Pawel Jakub Dawidek <pjd@FreeBSD.org>, phk@FreeBSD.org, Ivan Voras <ivoras@fer.hr>
Subject:   Re: Google SoC idea 
Message-ID:  <200506072225.aa16681@salmon.maths.tcd.ie>
In-Reply-To: Your message of "Tue, 07 Jun 2005 14:52:44 MDT." <42A6091C.40409@samsco.org> 

next in thread | previous in thread | raw e-mail | index | archive | help
> The problem with journalling at the block layer is that you pretty much 
> become forced to journal metadata and data, since the block layer really 
> doesn't know the distinction,

Definitely - I guess I should have stated that explicitly.

> Full journalling has many drawbacks from the viewpoint of 
> speed and complexity, of course.  So you really want to be able to do 
> just metadata journalling.

The complexity of full journaling (for filesystems that offer a
consistent version of themselves to the disk layer) would not seem
to be that high: a scheme like Ivan's seems to achieve it. Maybe
I've missed something? For us, journaling the metadata is of interest
because we're interested in avoiding fscks. I suppose full journaling
may be of interest to people with other applications, despite the
unfortunate performance in many situations.

> An alternate SoC project that would be very useful is block-level 
> snapshots.  I'm not sure if I'll be able to retain the filesystem 
> snapshot functionality in UFS with journalling enabled, so moving to 
> doing the snapshots in the block layer would be a good way to make up 
> for this.  Beware that while the GEOM transform would be pretty 
> straight-forward to write, the real trick comes from being able to make 
> the consumer of a block device (a filesystem, maybe) flush itself to a 
> consistent state while the snapshot is being taken.  The infrastructure 
> for this is the part that is very interesting, but also the most work.

Definitely another interesting project, though maybe a bit too much
work to ask someone to do for $4500 over a summer...

	David.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200506072225.aa16681>