From owner-freebsd-fs@FreeBSD.ORG Wed Mar 14 17:37:05 2012 Return-Path: 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 01D871065672 for ; Wed, 14 Mar 2012 17:37:05 +0000 (UTC) (envelope-from markm-lists@intellasoft.net) Received: from mail.mystoragebox.com (mail.mystoragebox.com [64.27.7.19]) by mx1.freebsd.org (Postfix) with ESMTP id D842E8FC15 for ; Wed, 14 Mar 2012 17:37:04 +0000 (UTC) Received: from cpe-67-240-79-210.nycap.res.rr.com ([67.240.79.210] helo=[192.168.35.110]) by mail.mystoragebox.com with esmtpsa (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.76) (envelope-from ) id 1S7roE-0002vk-E6 for freebsd-fs@freebsd.org; Wed, 14 Mar 2012 13:16:22 -0400 Message-ID: <4F60D73F.2010201@intellasoft.net> Date: Wed, 14 Mar 2012 13:37:03 -0400 From: Mark Murawski User-Agent: Mozilla/5.0 (X11; Linux i686; rv:10.0.2) Gecko/20120216 Thunderbird/10.0.2 MIME-Version: 1.0 To: freebsd-fs@freebsd.org References: <4F5F7116.3020400@intellasoft.net> <4F5F97A4.6070000@brockmann-consult.de> <4F60266D.1090302@intellasoft.net> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: ZFS file corruption problem X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 14 Mar 2012 17:37:05 -0000 On 03/14/12 10:03, Bob Friesenhahn wrote: > On Wed, 14 Mar 2012, Mark Murawski wrote: >> >> Why would the whole pool now become available upon access to a bad file? > > A disk drive (or HBA) may be hanging (e.g. endless retries) when the bad > file is accessed. This is a common problem with consumer disks or HBAs > which believe they are the top level authority when it comes to data > integrity. Zfs itself does not include any timers to decide to stop > waiting. Zfs depends on the lower-level OS & drivers to decide to stop > waiting on a stalled device. > > Bob Ah okay. And you assumed correctly that I had meant "become unavailable upon access". This is probably related to the bug I'm experiencing on my Promise SATA 300 TX4 controller (http://lists.freebsd.org/pipermail/freebsd-bugs/2012-March/047998.html)