Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 22 Sep 2006 09:44:28 +0200
From:      Ulf Lilleengen <lulf@stud.ntnu.no>
To:        freebsd-geom@freebsd.org
Subject:   Re: geom - help ...
Message-ID:  <20060922074428.GA1847@stud.ntnu.no>
In-Reply-To: <20060921200909.GA13927@megan.kiwi-computer.com>
References:  <45122531.6010503@infracaninophile.co.uk> <20060921093038.57966.qmail@web30303.mail.mud.yahoo.com> <20060921200909.GA13927@megan.kiwi-computer.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Sep 21, 2006 at 03:09:09PM -0500, Rick C. Petty wrote:
> On Thu, Sep 21, 2006 at 02:30:37AM -0700, R. B. Riddick wrote:
> 
> Is everybody forgetting about gvinum?  Just because it was totally broken
> in 6.0 and 5.5..
> 
> What are data centers using?  I can't imagine they're all doing gmirror and
> gstripe combinations..  gvinum gives you true volume management, e.g. it's
> easier to wipe a mirrored "volume" and make it raid5 or vice-versa.  It is
> useful to have concat + stripe + mirror + raid5 volumes managed together.
> 
> Recently I get the feeling that nobody cares about g/vinum anymore.  I'm
> all for different ways of doing the same thing, but please don't drop
> gvinum!  I certainly don't want to suffer through another bumpy transition,
> like 5.4 --> 6.1 was..
> 
> Since I'm on my pro-vinum rant:  People already have terabytes of data
> stored in vinum-managed volumes.  Forcing everybody to move things into the
> new geom containers won't make legacy users very happy.  I believe there
> are many 4.11 users who are waiting for gvinum to become as stable as vinum
> used to be.

Well, I can assure you that you're not alone :) I'm dedicating much of my
left-over time to work on gvinum, and you can see some of what I've been working
on in the last status report. Some of this will hopefully come into CURRENT soon
as Greg Lehey have had the time too review some patches. I'm almost done with
the concat command too, but everytime I'm working on gvinum i bump into many
things that could have been done better in terms of abstraction.

More on this can be found here:
http://lists.freebsd.org/pipermail/freebsd-arch/2006-August/005592.html

As mentioned later in that post, a new type of volume manager should be designed
to replace gvinum in the future. But it's still important to improve gvinum
because of that fact that many still uses it.

-- 
Ulf Lilleengen



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060922074428.GA1847>