From owner-freebsd-hardware@FreeBSD.ORG Wed Jul 18 02:21:10 2012 Return-Path: Delivered-To: freebsd-hardware@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 1895C1065678 for ; Wed, 18 Jul 2012 02:21:10 +0000 (UTC) (envelope-from lnb@freebsdsystems.com) Received: from panda.servaris.com (panda.servaris.com [107.6.50.5]) by mx1.freebsd.org (Postfix) with ESMTP id A64BA8FC14 for ; Wed, 18 Jul 2012 02:21:09 +0000 (UTC) Received: (qmail 68505 invoked by uid 89); 18 Jul 2012 02:21:08 -0000 Received: from unknown (HELO ?192.168.0.55?) (lnb@freebsdsystems.com@99.238.64.55) by panda.servaris.com with ESMTPA; 18 Jul 2012 02:21:08 -0000 Message-ID: <50061D65.7040605@freebsdsystems.com> Date: Tue, 17 Jul 2012 22:20:21 -0400 From: Lanny Baron Organization: Freedom Technologies Corp. FreeBSD Systems User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:13.0) Gecko/20120614 Thunderbird/13.0.1 MIME-Version: 1.0 To: freebsd-hardware@freebsd.org References: <201207170759.44995.erichfreebsdlist@ovitrap.com> In-Reply-To: Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: Server memory problems X-BeenThere: freebsd-hardware@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: General discussion of FreeBSD hardware List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 18 Jul 2012 02:21:10 -0000 Hi Andy, Sounds to me like you have 1) a flakey board, 2) the memory is not identical. We never use kingston for a variety of reasons, but you really should see if the part numbers are identical. The timings on the drams is critical. I would never mix capacities. 3) Make sure the memory is all the same i.e. registered e.c.c. or non registered e.c.c. I don't think its the power supply, but it can be. Regards, Lanny http://www.servaris.com or http://www.freebsdsystems.com On 7/17/2012 1:16 PM, Andy Young wrote: > Hi Erich, > > Why would the power supply be suspect since the machine is perfectly stable > with 64 GB of memory in it? > > The server won't stay up long enough to run memtest. > > Andy > > On Mon, Jul 16, 2012 at 8:59 PM, Erich Dollansky < > erichfreebsdlist@ovitrap.com> wrote: > >> Hi, >> >> On Tuesday 17 July 2012 06:45:18 Andy Young wrote: >>> I am having trouble with one of our servers and I'm not sure what to try >>> next. It has a Supermicro H8DGi-F motherboard with two 16-core AMD >>> processors and two memory banks, one for each processor. When I >> originally >>> built it, I only had one processor and 40 GB of ram. Everything worked >>> awesome. I recently upgraded it, adding another processor and another 40 >> GB >>> of ram. It was incredibly unstable and constantly rebooted within minute >> or >>> two of uptime, sometimes it wouldn't even boot all the way before >> crashing >>> and rebooting again. Seemed like a memory issue so I scaled it back to >> two >>> processors and 32 GB (4x8GB) of ram. Worked well so I added the >> remaining 8 >>> GB sticks I had, bringing it up to 64 GB. Still worked great. The sticks >> I >>> had left were a mix and match variety of 8GB and 4GB sticks. Thinking >> maybe >>> there was some problem with mixing them, I ordered more 8GB memory just >>> like the ones in the box. While waiting for the new memory, the machine >>> performed great with no issues. New memory arrived and I added two more >> 8GB >>> sticks. Immediately the constant crashing returned. It seems really >>> unlikely that I got bad memory in two separate orders. Does anyone have >> any >>> other ideas? Again, its perfectly stable with two processors and 64 GB of >>> memory but goes nuts when I more. >>> >> could it be caused by the power supply? >> >> Did you run a memory test? >> >> If possible, try different power supplies. >> >>> I really appreciate the help!! >>> >>> Motherboard: Supermicro H8DGi-F >>> CPU: 2 x AMD 6274 (2.2 Ghz 16-core) >>> Memory: Kingston 8GB DDR3 1333 >> >> No ECC? >> >> Erich >> > > >