From owner-freebsd-stable@FreeBSD.ORG Sun Mar 20 15:03:39 2005 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 94FF916A4CE; Sun, 20 Mar 2005 15:03:39 +0000 (GMT) Received: from athena.softcardsystems.com (mail.softcardsystems.com [12.34.136.114]) by mx1.FreeBSD.org (Postfix) with ESMTP id 13D6B43D46; Sun, 20 Mar 2005 15:03:39 +0000 (GMT) (envelope-from sah@softcardsystems.com) Received: from athena (athena [12.34.136.114])j2KEvVbL018830; Sun, 20 Mar 2005 09:57:32 -0500 Date: Sun, 20 Mar 2005 09:57:31 -0500 (EST) From: Sam X-X-Sender: sah@athena To: "Greg 'groggy' Lehey" In-Reply-To: <20050319231154.GA79501@wantadilla.lemis.com> Message-ID: References: <423CAE7E.4070009@dmv.com> <20050319231154.GA79501@wantadilla.lemis.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed cc: Sven Willenberger cc: freebsd-stable@freebsd.org Subject: Re: Creating a striped set of mirrors using gvinum X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 20 Mar 2005 15:03:39 -0000 > On Saturday, 19 March 2005 at 17:58:06 -0500, Sven Willenberger wrote: >> I am hoping someone has found a way to create this type of raid set >> using [g]vinum. I see that it is a trivial matter to create a mirror of >> 2 striped sets but I have not seen a way to create a stripe set out of >> multiple mirrored sets (e.g. stripe across 3 sets of mirrors). Has >> anyone managed to implement this and, if so, what does your >> configuration file look like? > > No, it doesn't work that way. > >> If not, could this be added as a feature request for gvinum? > > To do this you should enter a PR. It would help if you could explain > why it's preferable to mirrored plexes. It's preferable because: 1. After a single disk failure the rebuilding process only needs to rebuild one disk, not a whole stripe of disks. This also means that if your mirror code distributes reads to the mirrored disks for performance optimization, you don't lose that performance for the whole array when only one disk goes out of commission. 2. You can take multiple disk failures, one per mirror. In a mirror of stripes you can lose two before you're sunk -- the rebuild time of having to remirror an entire stripe on single failure means an even shakier bridge. This is to say that these are the benefits in other disk systems I've used. Spare management becomes more complex, unfortunately. sah