From owner-freebsd-current@FreeBSD.ORG  Tue Apr 10 21:49:02 2007
Return-Path: <owner-freebsd-current@FreeBSD.ORG>
X-Original-To: current@freebsd.org
Delivered-To: freebsd-current@FreeBSD.ORG
Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52])
	by hub.freebsd.org (Postfix) with ESMTP id 3402E16A404
	for <current@freebsd.org>; Tue, 10 Apr 2007 21:49:02 +0000 (UTC)
	(envelope-from craig@xfoil.gank.org)
Received: from ion.gank.org (ion.gank.org [69.55.238.164])
	by mx1.freebsd.org (Postfix) with ESMTP id 25F8D13C45A
	for <current@freebsd.org>; Tue, 10 Apr 2007 21:49:02 +0000 (UTC)
	(envelope-from craig@xfoil.gank.org)
Received: by ion.gank.org (Postfix, from userid 1001)
	id C5AE6110CF; Tue, 10 Apr 2007 16:30:46 -0500 (CDT)
Date: Tue, 10 Apr 2007 16:30:44 -0500
From: Craig Boston <craig@xfoil.gank.org>
To: current@freebsd.org
Message-ID: <20070410213044.GB19591@nowhere>
Mail-Followup-To: Craig Boston <craig@xfoil.gank.org>, current@freebsd.org
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.4.2.2i
Cc: 
Subject: Fun with ZFS
X-BeenThere: freebsd-current@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Discussions about the use of FreeBSD-current
	<freebsd-current.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-current>, 
	<mailto:freebsd-current-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-current>
List-Post: <mailto:freebsd-current@freebsd.org>
List-Help: <mailto:freebsd-current-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-current>,
	<mailto:freebsd-current-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Tue, 10 Apr 2007 21:49:02 -0000

Just playing around, doing some stress testing and see what I could
break.  Since zvols are exposed as GEOM providers I decided to see if it
would accept them in a pool.  Slightly surprising, it works without
incident:

# zfs create -V 4G tank/testvol
# zpool create test zvol/tank/testvol

The resulting /test seems full functional.  Though why anyone would want
to do such a thing I have no idea.

More in the realm of intentional foot-shooting, you can get a nice
deadlock by doing that recursively:

# zfs create -V 1G test/panicme
# zpool add test zvol/test/panicme

;)

Though as far as deadlocks go it's not a particularly bad one.  zpool
and zfs commands all hang in zfs:(& or zfs:&s states, but the
filesystems can still be accessed -- even the ones in the test pool
above.

Of course there can be no sane behavior to the above sequence of
commands, except perhaps a panic or a stern warning message if it's
really nice.

Next up: interactions with ggate and other GEOM classes.

Craig