From owner-freebsd-net@FreeBSD.ORG Sat Jan 16 13:29:48 2010 Return-Path: Delivered-To: freebsd-net@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 86B401065670 for ; Sat, 16 Jan 2010 13:29:48 +0000 (UTC) (envelope-from barney_cordoba@yahoo.com) Received: from web63903.mail.re1.yahoo.com (web63903.mail.re1.yahoo.com [69.147.97.118]) by mx1.freebsd.org (Postfix) with SMTP id 309618FC0C for ; Sat, 16 Jan 2010 13:29:47 +0000 (UTC) Received: (qmail 5534 invoked by uid 60001); 16 Jan 2010 13:29:45 -0000 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yahoo.com; s=s1024; t=1263648584; bh=OC5yIWTG3ZjIZI8mo6Ew6eXY+naOXFuu/s5NUPdgmP4=; h=Message-ID:X-YMail-OSG:Received:X-Mailer:Date:From:Subject:To:Cc:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=vyAk0rkDHxAqYdYe/JfWJlGh5mZdFJ4T2xTs8qemXLTsHDZ7dsxAGDznDiejMZNJ3Q+p2vVG0eE7gFx6Jq4y4LD9ThxR9UVEyCbN7q+m5pXkimKgdPk9Nr/jaKHnD1vaHcz88NVJGFBM8C32mHJRNCmS3RDMLtSQOL03jIg0fqI= DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=Message-ID:X-YMail-OSG:Received:X-Mailer:Date:From:Subject:To:Cc:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=6fsJusHTHeuWGZEeEeYzO8PPWO7VxL3AylHE0frS/C178JC9Vbnf+yIMYkP+jDIgCgQfbP3Pj65JBVNkZfFX3cypxMKF3CYlryc/jLRach3/Te83o7v+ytjhFTzgRzhAuw4vOEG83DszWbGb0n8UogKVnWQuHC2MxpEnpzUtvR4=; Message-ID: <965936.5529.qm@web63903.mail.re1.yahoo.com> X-YMail-OSG: VW5dx04VM1mG7AF6mpfDlU_6VVUQJ6b4hsHy19.Ivz3irP33BBHyH7cD5vvUksJWjo69WVun8DI3UlES._Uli_tSxf2WqMFjfhILKDInGa4L_m.YTNK89wp3JYHCRc4mWZ1DcALs.WHYg44aq2vklREQz92Nmvg3i70zrtRtN25NY9cz3PI6HjM0ovuMFsQtRQxXohp.aRPYSz0RY_YqOfJ5YgItjigV1EdSWna5C9cMDYBk_5hGtrYHWf4MXrDXBHyyJwCWgikWCx89nRM9to3MAsvA1M75qzd0ZaU- Received: from [98.203.21.152] by web63903.mail.re1.yahoo.com via HTTP; Sat, 16 Jan 2010 05:29:44 PST X-Mailer: YahooMailClassic/9.0.20 YahooMailWebService/0.8.100.260964 Date: Sat, 16 Jan 2010 05:29:44 -0800 (PST) From: Barney Cordoba To: Sebastian Hyrwall In-Reply-To: <4B504290.5020506@keff.org> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Cc: freebsd-net@freebsd.org Subject: Re: FreeBSD 7.2 vs Linux in routing performance X-BeenThere: freebsd-net@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Networking and TCP/IP with FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 16 Jan 2010 13:29:48 -0000 =0A=0A--- On Fri, 1/15/10, Sebastian Hyrwall wrote:=0A=0A> Fr= om: Sebastian Hyrwall =0A> Subject: FreeBSD 7.2 vs Linux in ro= uting performance=0A> To: freebsd-net@freebsd.org=0A> Date: Friday, January= 15, 2010, 5:25 AM=0A> Hi=0A> =0A> I have to identical x86 routers with the= following=0A> specifications,=0A> =0A> hw.model: Intel(R) Atom(TM) CPU=A0= =0A> 330=A0=A0=A0@ 1.60GHz=0A> hw.physmem: 2132996096=0A> hw.usermem: 17872= 52736=0A> hw.realmem: 2146041856=0A> 2x re0: 8168/8168B/8168C/= 8168CP/8168D/8111B/8111C/8111CP PCIe=0A> Gigabit Ethernet> port 0xd800-0xd8= ff mem=0A> 0xfeaff000-0xfeafffff,0xfdef0000-0xfdefffff irq 16 at device=0A>= 0.0 on pci2=0A> =0A> I know it's not really the best equipment to use in= =0A> gbit-enviroments but that is irrelevant here.=0A> =0A> One of these ru= ns FreeBSD 7.2 (R-p4) and the other Linux=0A> 2.6.31.5.=0A> =0A> Without pf= /iptables loaded the FreeBSD-server maxes out at=0A> 35MB/s when it comes t= o forwarding between the two NICs=0A> (simple http-transfer used for testin= g).=0A> The Linux-server pushes 90-100MB/s between the NICs with=0A> the sa= me test. Both servers are connected the same way to=0A> the network (I swap= them between the testing).=0A> =0A> Any suggestions on where the gigantic = performance loss=0A> might be and how to fix it?=0A> =0A> I intend to switc= h FreeBSD 8 in the coming month and maybe=0A> that will fix the problem but= I am hoping it's also fixable=0A> in 7.2.=0A> =0A=0AThe equipment really I= S relevant. The FreeBSD realtek driver is=0Aparticularly sucky. As with any= free OS, some drivers are good, =0Aand most are not. So try something else= . Anything written by=0ABill Paul is assembly-line quality by definition.= =0A=0A=0ABarney=0A=0A=0A