From owner-freebsd-proliant@FreeBSD.ORG Fri Apr 18 09:37:45 2008 Return-Path: Delivered-To: freebsd-proliant@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id A1DFD1065673 for ; Fri, 18 Apr 2008 09:37:45 +0000 (UTC) (envelope-from jdc@parodius.com) Received: from mx01.sc1.parodius.com (mx01.sc1.parodius.com [72.20.106.3]) by mx1.freebsd.org (Postfix) with ESMTP id 780FE8FC24 for ; Fri, 18 Apr 2008 09:37:45 +0000 (UTC) (envelope-from jdc@parodius.com) Received: by mx01.sc1.parodius.com (Postfix, from userid 1000) id 2906C1CC033; Fri, 18 Apr 2008 02:20:34 -0700 (PDT) Date: Fri, 18 Apr 2008 02:20:34 -0700 From: Jeremy Chadwick To: Gianni Message-ID: <20080418092034.GA34069@eos.sc1.parodius.com> References: <20080417171439.M22551@bellera.cat> <20080417172729.M94092@bellera.cat> <83ABD599-A699-4CD8-AF1F-D42357C7D9B9@yahoo.it> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <83ABD599-A699-4CD8-AF1F-D42357C7D9B9@yahoo.it> User-Agent: Mutt/1.5.17 (2007-11-01) Cc: freebsd-stable@freebsd.org, freebsd-proliant@freebsd.org Subject: Re: gmirror problem with HP Proliant ML110 G5 X-BeenThere: freebsd-proliant@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: "Technical discussion of FreeBSD on HP ProLiant server platforms." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 18 Apr 2008 09:37:45 -0000 On Fri, Apr 18, 2008 at 07:12:06AM +0200, Gianni wrote: > On 17/apr/08, at 19:28, Josep Pujadas i Jubany wrote: >> Two weeks ago I installed FreeBSD 7.0 in a new HP Proliant ML110 G5 >> machine >> and I configured ad0 for using gmirror, waiting for a second disk. >> >> I just added a second hard disk arrived today from HP ... >> >> # gmirror insert gm0 /dev/ad2 >> >> I'm having DMA errors: >> >> Apr 17 16:49:55 mail_2 kernel: GEOM_MIRROR: Device gm0: rebuilding >> provider >> ad2. >> Apr 17 16:50:13 mail_2 kernel: ad2: TIMEOUT - WRITE_DMA retrying (1 retry >> left) LBA=1534720 >> Apr 17 16:50:46 mail_2 kernel: ad2: TIMEOUT - WRITE_DMA retrying (1 retry >> left) LBA=4563840 >> ... >> ... >> Apr 17 17:39:28 mail_2 kernel: ad2: TIMEOUT - WRITE_DMA retrying (1 retry >> left) LBA=268281088 >> Apr 17 17:39:46 mail_2 kernel: ad2: TIMEOUT - WRITE_DMA48 retrying (1 >> retry >> left) LBA=269601536 >> Apr 17 17:39:46 mail_2 kernel: ad2: FAILURE - WRITE_DMA48 >> status=51 error=10 LBA=269601536 >> Apr 17 17:39:46 mail_2 kernel: GEOM_MIRROR: Synchronization request failed >> (error=5). ad2[WRITE(offset=138035986432, length=131072)] >> Apr 17 17:39:46 mail_2 kernel: GEOM_MIRROR: Device gm0: provider ad2 >> disconnected. >> Apr 17 17:39:46 mail_2 kernel: GEOM_MIRROR: Device gm0: rebuilding >> provider >> ad2 stopped. >> >> Disks are equal in capactiy: >> >> Apr 17 16:40:48 mail_2 kernel: ad0: 238475MB at >> ata0-master SATA150 >> Apr 17 16:40:48 mail_2 kernel: ad2: 238475MB at ata1- >> master SATA150 >> >> # atacontrol list >> >> ATA channel 0: >> Master: ad0 Serial ATA v1.0 >> Slave: no device present >> ATA channel 1: >> Master: ad2 Serial ATA v1.0 >> Slave: no device present >> ATA channel 2: >> Master: acd0 Serial ATA v1.0 >> Slave: no device present >> ATA channel 3: >> Master: no device present >> Slave: no device present >> >> Any ideas? Thanks, >> >> Josep Pujadas Josep, the disks may be the same in capacity, but they aren't completely identical. It's fairly obvious one is a Seagate and the other is HP/Compaq drive. This is very likely **not** the cause of the DMA errors you're seeing, but I did want to take a moment to state that mix-matching drives with different semantics in a mirror is somewhat risky. > There are a few reports of this on the freebsd-stable mailing list in March > and I'm also experiencing the same problem, so far none of the suggested > resolutions has helped for me. > http://lists.freebsd.org/pipermail/freebsd-stable/2008-March/thread.html > > I've got the issue on a custom built box with amd64 + 7.0 + gmirror and > I've got a Proliant ML110 G5 running 6.3 I'd like to upgrade to 7.0 but too > scared to do so until this issue appears to have been identified and fixed. > Copying in freebsd-stable as it's not just a problem on Proliant. > -Gianni I've documented the DMA problem quite thoroughly. The DMA errors are not specific to gmirror: http://wiki.freebsd.org/JeremyChadwick/Commonly_reported_issues If the problem is easily repeatable, and you have serial console available on the box, please contact Scott Long who has offered to help track the source of these problems down. -- | Jeremy Chadwick jdc at parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB |