From owner-freebsd-fs@FreeBSD.ORG  Wed Nov 17 02:13:00 2010
Return-Path: <owner-freebsd-fs@FreeBSD.ORG>
Delivered-To: freebsd-fs@freebsd.org
Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34])
	by hub.freebsd.org (Postfix) with ESMTP id 6F74E1065672;
	Wed, 17 Nov 2010 02:13:00 +0000 (UTC) (envelope-from TERRY@tmk.com)
Received: from server.tmk.com (server.tmk.com [204.141.35.63])
	by mx1.freebsd.org (Postfix) with ESMTP id 488438FC15;
	Wed, 17 Nov 2010 02:13:00 +0000 (UTC)
Received: from tmk.com by tmk.com (PMDF V6.4 #37010)
	id <01NUC5J6QBZK00BNN4@tmk.com>; Tue, 16 Nov 2010 21:12:57 -0500 (EST)
Date: Tue, 16 Nov 2010 20:41:47 -0500 (EST)
From: Terry Kennedy <TERRY@tmk.com>
In-reply-to: "Your message dated Mon, 15 Nov 2010 23:48:26 -0800"
	<816D59CD-1BAF-4331-BEAD-67CEADCE4EF9@deman.com>
To: Michael DeMan <freebsd@deman.com>
Message-id: <01NUC6V4LBAQ00BNN4@tmk.com>
MIME-version: 1.0
Content-type: TEXT/PLAIN; CHARSET=us-ascii
References: <01NUB1F8POL000BNN4@tmk.com> <01NUB3IOMZJW00BNN4@tmk.com>
Cc: freebsd-fs@freebsd.org, freebsd-stable@freebsd.org
Subject: Re: ZFS panic after replacing log device
X-BeenThere: freebsd-fs@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Filesystems <freebsd-fs.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-fs>,
	<mailto:freebsd-fs-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-fs>
List-Post: <mailto:freebsd-fs@freebsd.org>
List-Help: <mailto:freebsd-fs-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-fs>,
	<mailto:freebsd-fs-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Wed, 17 Nov 2010 02:13:00 -0000

> I would say it is definitely very odd that writes are a problem.  Sounds
> like it might be a hardware problem.  Is it possible to export the pool, 
> remove the ZIL and re-import it?  I myself would be pretty nervous trying
> that, but it would help isolate the problem?  If you can risk it.

  I think it is unlikely to be a hardware problem. While I haven't run any
destructive testing on the ZFS pool, the fact that it can be read without
error, combined with ECC throughout the system and the panic always happen-
ing on the first write, makes me think that it is a software issue in ZFS.

  When I do:

	zpool export data; zpool remove data da0

  I get a "No such pool: data". I then re-imported the pool and did:

	zpool offline data da0; zpool export data; zpool import data

  After doing that, I can write to the pool without a panic. But once I
online the log device and do any writes, I get the panic again.

  As I mentioned, I have this data replicated elsewere, so I can exper-
iment with the pool if it will help track down this issue.

        Terry Kennedy             http://www.tmk.com
        terry@tmk.com             New York, NY USA