Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 15 Sep 2006 11:45:39 +0200 (CEST)
From:      Oliver Fromme <olli@lurza.secnetix.de>
To:        freebsd-geom@FreeBSD.ORG
Subject:   Re: gmirror size question
Message-ID:  <200609150945.k8F9jdU6007297@lurza.secnetix.de>
In-Reply-To: <20060915090514.90093.qmail@web30314.mail.mud.yahoo.com>

next in thread | previous in thread | raw e-mail | index | archive | help
R. B. Riddick <arne_woerner@yahoo.com> wrote:
 > Oliver Fromme <olli@lurza.secnetix.de> wrote:
 > > PS:  I think that hint should be mentioned in the Handbook
 > > section about gmirror.  It could prevent major foot-shoting.
 > > Comments?
 > > 
 > Hmm...
 > 
 > I would say, if one buys a new disk, that is 1 sector too short *giggle*, one
 > could still try the following plan:
 > 0. single user mode
 > 1. Make a new file system on the shorter&new disk
 > 2. Move the data from the old disk to the new disk (MBR with dd, the rest with
 > cpio or so)
 > 3. Create a new gmirror just with the new disk as the only disk
 > 4. Add the old disk to the new gmirror
 > 5. Change the configuration in order to reflect the changed name of the
 > gmirror.
 > QEF (reboot inbetween whenever it seems necessary)

Yes, but that takes quite a lot of work and time.  If you
use RAID1, you don't want any downtime in the first place.
Also, it might not be obvious to the person repairing the
mirror that the new disk is too small.  The EINVAL errno
code isn't really very verbose about the root of the
problem.

If you use the gnop trick at the beginning, so the gmirror
is somewhat smaller, then you can simply plug in the new
drive, gmirror insert, and you're done.  No downtime at
all.  You don't even have to remember anything special.

Best regards
   Oliver

-- 
Oliver Fromme,  secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing
Dienstleistungen mit Schwerpunkt FreeBSD: http://www.secnetix.de/bsd
Any opinions expressed in this message may be personal to the author
and may not necessarily reflect the opinions of secnetix in any way.

"[...]  one observation we can make here is that Python makes
an excellent pseudocoding language, with the wonderful attribute
that it can actually be executed."  --  Bruce Eckel



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200609150945.k8F9jdU6007297>