From owner-freebsd-chat Sat Jan 6 21: 8:47 2001 Delivered-To: freebsd-chat@freebsd.org Received: from grumpy.dyndns.org (user-24-214-56-41.knology.net [24.214.56.41]) by hub.freebsd.org (Postfix) with ESMTP id 970A437B400 for ; Sat, 6 Jan 2001 21:08:29 -0800 (PST) Received: from localhost (localhost [127.0.0.1]) by grumpy.dyndns.org (8.11.1/8.11.1) with ESMTP id f0756as07326; Sat, 6 Jan 2001 23:06:36 -0600 (CST) (envelope-from dkelly@grumpy.dyndns.org) Message-Id: <200101070506.f0756as07326@grumpy.dyndns.org> X-Mailer: exmh version 2.2 06/23/2000 with nmh-1.0.4 To: Francisco Reyes Cc: FreeBSD Chat List From: David Kelly Subject: Re: ECC worth the extra cost for SOHO server? In-reply-to: Message from Francisco Reyes of "Sat, 06 Jan 2001 15:16:03 EST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 06 Jan 2001 23:06:36 -0600 Sender: owner-freebsd-chat@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Francisco Reyes writes: > Is ECC woth the money for a home server? > It will be running a few lightly visited web sites. > I understand what ECC does, but what I don't have a clue is how often > errors would occur under normal circustance which would be trapped by ECC. > > As long as I keep the case cooled properly why would there be memory > errors? > > Any recommendation on ECC motherboards? So far I have only seen some Tyan > mobos listed with ECC. I recommend ECC for everyone. The memory should be no more than 12% more expensive. The real use of ECC isn't so much the real time correction of errors, but for the real time *detection* of errors. Otherwise it can be very difficult to track down the difference between a random software error and a memory error. Unless a memory chip fails catastrophically, its hard to find the errant chip. When they go bad they usually start logging an ECC correction every day or two. Rarely do they get into perpetual ECC corrections (as would indicate a catastrophic failure). Used to have about 25 Sun and SGI machines. Over a 3 year period we had about one memory SIMM replaced every 3 months. Must be considered that we had more memory in these machines than was common in PC's in 1996. Then again, they were better built than PC hardware. Over the last 12 years I have had memory fail in my own computers 3 times. Not counting new memory that arrived DOA (duh! don't ship it in white peanuts, use pink ones if you must). As for MBs? The new VIA KT133 does *not* do ECC (Asus A7V) but am told the older KX133 does. The Intel chipsets for Pentium-II and up do, but I don't know about the low cost chipsets such as 810. The good old BX chipset (and Asus P3B-F) certainly does. With all that said, I don't know what FreeBSD does in the face of an ECC correction event. Not sure if it logs or not. -- David Kelly N4HHE, dkelly@hiwaay.net ===================================================================== The human mind ordinarily operates at only ten percent of its capacity -- the rest is overhead for the operating system. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message