From owner-freebsd-stable@FreeBSD.ORG Mon Nov 28 20:55:11 2005 Return-Path: X-Original-To: stable@freebsd.org Delivered-To: freebsd-stable@FreeBSD.ORG Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0000016A41F for ; Mon, 28 Nov 2005 20:55:10 +0000 (GMT) (envelope-from ltning@anduin.net) Received: from anduin.net (anduin.net [212.12.46.226]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5959F43D58 for ; Mon, 28 Nov 2005 20:55:10 +0000 (GMT) (envelope-from ltning@anduin.net) Received: from ranger.anduin.net ([81.0.162.52] helo=[192.168.1.112]) by anduin.net with esmtpa (Exim 4.50 (FreeBSD)) id 1Egq1d-0001rV-RP; Mon, 28 Nov 2005 21:55:01 +0100 In-Reply-To: <84dead720511280654j138635abgcb9cc0978e6c26b7@mail.gmail.com> References: <93F6B911-8C64-4F5C-81F9-80EC271ED298@anduin.net> <84dead720511280545v2bc0bc35jd107da06b9a788cb@mail.gmail.com> <20187843-76FC-4EAB-AFF8-7493FB0C0077@anduin.net> <84dead720511280654j138635abgcb9cc0978e6c26b7@mail.gmail.com> Mime-Version: 1.0 (Apple Message framework v746.2) Content-Type: text/plain; charset=ISO-8859-1; delsp=yes; format=flowed Message-Id: <02757598-222D-408E-8B33-C2EE1E6E426E@anduin.net> Content-Transfer-Encoding: quoted-printable From: =?ISO-8859-1?Q?Eirik_=D8verby?= Date: Mon, 28 Nov 2005 21:54:30 +0100 To: Joseph Koshy X-Mailer: Apple Mail (2.746.2) Cc: stable@freebsd.org Subject: Re: Reduced java/tomcat performance 6-beta3 -> 6-stable ? X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 28 Nov 2005 20:55:11 -0000 Hi, I think I have found the culprit. There must be some sort of =20 difference between the machines after all (BIOS revision?), because =20 while on one machine the interrupt rate for the bge card stays very =20 low (2 to be exact) during maximum load, the other machine goes =20 beyond 1000 and keeps rising constantly. This might also explain why =20 performance slowly degrades over time on that machine, and response =20 times vary wildly, while the "fast" machine responds nicely within =20 1-2 seconds no matter the load and testing time. I will have to investigate this more closely. Is there a way to force =20= the NIC to polling mode (I'm assuming that is the difference, an IRQ =20 rate of 2 is too low for a heavily loaded server if the NIC is =20 interrupt-driven)? Anything else I could look at? Also, the interrupt rates for the CPUs stay at 2000 sharp on the fast =20= system, but fluctuates somewhat on the other. /Eirik On Nov 28, 2005, at 15:54 , Joseph Koshy wrote: > E=D8> *loads* more context switches than on the BETA-3 system. > E=D8> I have not yet tried this during load > > - Which scheduler have you configured (BSD or ULE)? > - What do the interrupt statistics show? Any interrupt > storms? Please check the mailing lists for a prior > discussion on interrupt storms on some motherboards. > - Could you post the dmesg output from the systems (I > presume there aren't any significant differences). > > Please CC -stable too. > > -- > FreeBSD Volunteer, http://people.freebsd.org/~jkoshy > >