From owner-freebsd-smp@FreeBSD.ORG Fri Jan 7 00:13:42 2005 Return-Path: Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3BBE416A4CE; Fri, 7 Jan 2005 00:13:42 +0000 (GMT) Received: from stewie.obfuscated.net (stewie.obfuscated.net [66.118.188.125]) by mx1.FreeBSD.org (Postfix) with ESMTP id 00CF143D2F; Fri, 7 Jan 2005 00:13:42 +0000 (GMT) (envelope-from m@obmail.net) Received: from [192.168.1.104] (653259hfc120.tampabay.rr.com [65.32.59.120]) (using TLSv1 with cipher RC4-SHA (128/128 bits)) (No client certificate requested) by stewie.obfuscated.net (Postfix) with ESMTP id 31A1D6104; Thu, 6 Jan 2005 19:13:41 -0500 (EST) In-Reply-To: References: Mime-Version: 1.0 (Apple Message framework v619) Content-Type: text/plain; charset=US-ASCII; format=flowed Message-Id: Content-Transfer-Encoding: 7bit From: M Date: Thu, 6 Jan 2005 19:12:53 -0500 To: Robert Watson X-Mailer: Apple Mail (2.619) cc: freebsd-smp@freebsd.org Subject: Re: Dual Xeon 2.8 with HT - How is the support in FreeBSD 5.3 X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 07 Jan 2005 00:13:42 -0000 On Jan 6, 2005, at 6:14 PM, Robert Watson wrote: > On Thu, 6 Jan 2005, M wrote: > >> The only issue I've had on any of them is a highly loaded NFS server >> which has crashed, but > > If you have any further debugging information on the NFS crash, I'd be > happy to take a look. It could well have been fixed by the recent > RELENG_5_3 NFS fix, which has been in RELENG_5 for a month or two. > Robert, This morning I applied p3 and it's running, we'll see how long it goes. I turned the access_cache_timeout on the 10 client systems down to 0. I'm pushing about 35 Mbit/sec of traffic both ways at the moment. At this point the clients are generating nfs server not responding/nfs server is alive again messages every few minutes so if p3 doesn't fix the issue I should see something. (I don't think the error isn't coming from the disks. The FC disk array has plenty of spare capacity as measured through gstat's %busy, mostly from having 4 GB of memory, most of it Inactive). Unfortunately the crashes produce no output at all, just hard hang. If it continues I'm going to turn off SMP and go single processor. I understand that may help get some useful output from a system crash. Do you have any thoughts on hyperthreading on a system that's used only as an NFS server? Thanks