From owner-freebsd-questions@FreeBSD.ORG Wed Jun 22 14:49:02 2005 Return-Path: X-Original-To: freebsd-questions@freebsd.org Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 4683916A41C for ; Wed, 22 Jun 2005 14:49:02 +0000 (GMT) (envelope-from ean@hedron.org) Received: from prosporo.hedron.org (hedron.org [66.11.182.60]) by mx1.FreeBSD.org (Postfix) with ESMTP id 784C243D48 for ; Wed, 22 Jun 2005 14:49:01 +0000 (GMT) (envelope-from ean@hedron.org) Received: from localhost.hedron.org (localhost.hedron.org [127.0.0.1]) by prosporo.hedron.org (Postfix) with ESMTP id 022F3C2DF for ; Wed, 22 Jun 2005 10:49:13 -0400 (EDT) From: Ean Kingston To: freebsd-questions@freebsd.org Date: Wed, 22 Jun 2005 10:49:12 -0400 User-Agent: KMail/1.8 References: <20050622064024.GA19456@lothlorien.nagual.st> In-Reply-To: <20050622064024.GA19456@lothlorien.nagual.st> X-Face: W{mkf[fd1042ubL1FZ(CABIMzn~rdu<:SW\^LF_RB' Subject: Re: cloning with nfs? X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 22 Jun 2005 14:49:02 -0000 On June 22, 2005 02:40 am, Dick Hoogendijk wrote: > Yesterday I ruined my partition table on one of my machines. > Luckely this machine was almost an exact copy of another that still is > running fine. > > So, I can follow the procedure of copying one disk to another (following > the handbook). But this requires a fysical removal / action on the > machines and harddisks witch I don't want to do if not needed. > > I did a minimal install on the crashed machine (#B) > If disk'cloning' can be done through NFS that'll be the way to go for > me. > Will it be enough to export /var /usr /tmp and / (#B) to mountpoints on > machine #A and then follow the 'normal' dump/restore procedure mentioned > in the handbook? > Or are there side_effects and will fysical placement of the 'new' drive > in machine #A be the right way to do it? I don't think restore works reliably on NFS mounted disks but I have copied disks using dump/restore through ssh. I would not do a blind dump/restore of / or /var. Those filesystems can contain some installation specific information. I think the only thing out of / that you need to copy would be /etc and possibly /boot if you have a custom kernel. Just remember that a kernel install is not as simple as copying files. You don't need to copy /tmp since it should not contain any information that is needed to survive a reboot. Just reboot after you restore. As for /usr you should be able to dump/restore that one. If you have additional packages installed, you will also want to copy /var/db/pkg and possibly /var/db/ports. Likewise, if the system is a mail server, you will want to copy over the appropriate directory structure (typically /var/spool) but you need to make sure you don't copy over any of the spool files or your users are going to get 2 copies of the same message delivered. > Thanks for any advice. -- Ean Kingston E-Mail: ean AT hedron DOT org URL: http://www.hedron.org/ I am currently looking for work. If you need competent system/network administration please feel free to contact me directly.