From owner-freebsd-hackers@FreeBSD.ORG Sun Nov 9 14:07:09 2003 Return-Path: Delivered-To: freebsd-hackers@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id EF40416A4CF; Sun, 9 Nov 2003 14:07:09 -0800 (PST) Received: from VARK.homeunix.com (adsl-68-123-40-77.dsl.pltn13.pacbell.net [68.123.40.77]) by mx1.FreeBSD.org (Postfix) with ESMTP id ED18E43FDD; Sun, 9 Nov 2003 14:07:08 -0800 (PST) (envelope-from das@FreeBSD.ORG) Received: from VARK.homeunix.com (localhost [127.0.0.1]) by VARK.homeunix.com (8.12.9/8.12.9) with ESMTP id hA9M5gen014255; Sun, 9 Nov 2003 14:05:42 -0800 (PST) (envelope-from das@FreeBSD.ORG) Received: (from das@localhost) by VARK.homeunix.com (8.12.9/8.12.9/Submit) id hA9M5gKK014254; Sun, 9 Nov 2003 14:05:42 -0800 (PST) (envelope-from das@FreeBSD.ORG) Date: Sun, 9 Nov 2003 14:05:42 -0800 From: David Schultz To: Lukas Ertl Message-ID: <20031109220542.GA14205@VARK.homeunix.com> Mail-Followup-To: Lukas Ertl , Christian Laursen , hackers@FreeBSD.ORG, phk@FreeBSD.ORG References: <20031109003617.K626@korben.in.tern> <20031109023041.GA9171@VARK.homeunix.com> <868ympydcx.fsf@borg.borderworlds.dk> <20031109213450.GB13982@VARK.homeunix.com> <20031109224505.E58774@korben.in.tern> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20031109224505.E58774@korben.in.tern> cc: hackers@FreeBSD.ORG cc: Christian Laursen cc: phk@FreeBSD.ORG Subject: Re: geom_mirror implementation X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 09 Nov 2003 22:07:10 -0000 On Sun, Nov 09, 2003, Lukas Ertl wrote: > > If there's a good reason ccd(4) is harder to fix than geom_mirror, then > > you might want to talk to phk about rewriting geom_ccd based on > > geom_mirror. I believe scottl and phk have plans to fix raidframe, > > though, which would address a lot of the present limitations. > > I haven't used ccd or raidframe before, and vinum is still a lot of > 'voodoo'. I just wanted to play around and do something useful; > furthermore I learned a real lot about GEOM internals. If you're feeling up to it, I think people would much appreciate some of the problems in ccd being fixed. One of the things that ccd needs and that is probably pretty straightforward is a better algorithm to select which side of the mirror to read from. The present algorithm is broken and generally winds up with one of the disks satisfying all the requests unless a read error is reported.