From owner-freebsd-current@FreeBSD.ORG Wed Feb 4 07:27:24 2009 Return-Path: Delivered-To: current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 99DA31065689 for ; Wed, 4 Feb 2009 07:27:24 +0000 (UTC) (envelope-from nakal@web.de) Received: from fmmailgate02.web.de (fmmailgate02.web.de [217.72.192.227]) by mx1.freebsd.org (Postfix) with ESMTP id 281B88FC0A for ; Wed, 4 Feb 2009 07:27:23 +0000 (UTC) (envelope-from nakal@web.de) Received: from smtp06.web.de (fmsmtp06.dlan.cinetic.de [172.20.5.172]) by fmmailgate02.web.de (Postfix) with ESMTP id B03A5F9E1E26; Wed, 4 Feb 2009 08:27:22 +0100 (CET) Received: from [217.236.23.4] (helo=zelda.local) by smtp06.web.de with asmtp (TLSv1:AES128-SHA:128) (WEB.DE 4.110 #277) id 1LUcAM-00076D-00; Wed, 04 Feb 2009 08:27:22 +0100 Date: Wed, 4 Feb 2009 08:27:18 +0100 From: Martin To: Marcel Moolenaar Message-ID: <20090204082718.0f217b1a@zelda.local> In-Reply-To: References: <20090203082153.565746e2@zelda.local> <20090204004534.11ccca19@zelda.local> X-Mailer: Claws Mail 3.7.0 (GTK+ 2.14.7; amd64-portbld-freebsd8.0) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: nakal@web.de X-Sender: nakal@web.de X-Provags-ID: V01U2FsdGVkX1/ac6xM1pP6ok3IgYzi9+eNfH2HoVlXsVDpRjVZ wQpC5yaqibU4DAin/qbhPFfywdAYg4EEVOOEK+qkTAg5HjWEW5 +AqYTLzMs= Cc: FreeBSD Current Subject: Re: GEOM_PART: a quick update on logical partitions X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 04 Feb 2009 07:27:25 -0000 Am Tue, 03 Feb 2009 16:30:14 -0800 schrieb Marcel Moolenaar : > > Then, it is quite unusual to insert a new logical partition and copy > > all entries in the list one place down. > > What do you mean copy? I try to figure out why you think that the device names will change, when you simply use the offset in the partition list. When you want to insert a new partition at the beginning and you already have one at list offset 0 then you have to copy it one place down to 1. I thought that you want to offer a special solution for this problem. > Sure, but you can remove any logical partition in any > order, which means that subsequent adds can also be > in any order... Ok, so when you have logical partition 0, 1 and 2. /dev/ad0s1.0 /dev/ad0s1.1 /dev/ad0s1.2 you can remove partitions 0 and 1, and you get: /dev/ad0s1.2 now you can insert 0 again and you get: /dev/ad0s1.0 /dev/ad0s1.2 You can still use softlinks here. In the last situation you would get ad0s5 and ad0s7. Do I forget about something? > Yes, but those will be using the "fixed" names, not the > softlinks. I see. But you have to be aware that you create a special case here. Someone who starts "geli journal ad0s5" won't get "ad0s5.journal". And is it really a good idea to have the partition block number for the suffix? Imagine you use gpart or partition magic to move all partitions 1GB down, for example to enlarge the file system on the first primary partition. You would need to figure out each logical partition block number of the extended partition to fix the situation, if you use geom_eli/geom_journal before mounting. You would have to edit it in rc.conf for geli devices that appear on boot and fstab for for both, journal and eli. When you don't touch the logical layout itself, it shouldn't change the device names, in my opinion. -- Martin