From owner-freebsd-net@FreeBSD.ORG Sat Oct 21 20:12:59 2006 Return-Path: X-Original-To: freebsd-net@freebsd.org Delivered-To: freebsd-net@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id EE0DC16A412 for ; Sat, 21 Oct 2006 20:12:59 +0000 (UTC) (envelope-from lists@codeangels.com) Received: from mail.codeangels.com (monkey.codeangels.com [62.2.169.19]) by mx1.FreeBSD.org (Postfix) with ESMTP id 95E9C43D55 for ; Sat, 21 Oct 2006 20:12:57 +0000 (GMT) (envelope-from lists@codeangels.com) Received: (qmail-ldap/ctrl 13585 invoked from network); 21 Oct 2006 20:12:55 -0000 Received: from monkey.codeangels.com (HELO www.codeangels.com) (jsmcud@[192.168.5.6]) (envelope-sender ) by monkey.codeangels.com (qmail-ldap-1.03) with SMTP for ; 21 Oct 2006 20:12:55 -0000 Message-ID: <1355.192.168.1.6.1161461575.squirrel@www.codeangels.com> In-Reply-To: References: <2108.192.168.1.6.1161393460.squirrel@www.codeangels.com> Date: Sat, 21 Oct 2006 22:12:55 +0200 (CEST) From: "Kirill Ponazdyr" To: freebsd-net@freebsd.org User-Agent: SquirrelMail/Codeangels_GEN MIME-Version: 1.0 Content-Type: text/plain;charset=iso-8859-1 X-Priority: 3 Importance: Normal Subject: Re: Gigabit performance test X-BeenThere: freebsd-net@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: lists@codeangels.com List-Id: Networking and TCP/IP with FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 21 Oct 2006 20:13:00 -0000 > Are you using amd64 or i386 kernel ? the config implies you are using > i386 On firewall, i386 > > I found > kern.polling.idle_poll=1 > to improve performance in polling. Oh, that does it! Excellent! Performance jumped to 930mbit in single thread instantly. > Also, try updating the box to 6.2 > first as there are quite a few improvements to the em driver Done, it seems like CPU usage is less now > You can also fiddle with assigning less to userspace with > kern.polling.user_frac=30 > Don remember for sure, but ipfw seemed to be a bit faster that pf. > Also with no firewall loaded there seemed to be quite a bit more > throughput... However, that kind of defeats the purpose of a firewall. ;) Now the actual firewall test comes. pv vs ipf vs ipfw vs SonicWall 5060. Stay tuned for results. Thanks! Kirill