From owner-freebsd-net@FreeBSD.ORG Mon Dec 20 20:42:49 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 B66CA106566C for ; Mon, 20 Dec 2010 20:42:49 +0000 (UTC) (envelope-from lev@serebryakov.spb.ru) Received: from ftp.translate.ru (ftp.translate.ru [80.249.188.42]) by mx1.freebsd.org (Postfix) with ESMTP id 730D38FC15 for ; Mon, 20 Dec 2010 20:42:49 +0000 (UTC) Received: from lion.home.serebryakov.spb.ru (89.112.15.178.pppoe.eltel.net [89.112.15.178]) (Authenticated sender: lev@serebryakov.spb.ru) by ftp.translate.ru (Postfix) with ESMTPA id 3AD9313DF5F; Mon, 20 Dec 2010 23:42:48 +0300 (MSK) Date: Mon, 20 Dec 2010 23:42:41 +0300 From: Lev Serebryakov X-Priority: 3 (Normal) Message-ID: <1715104467.20101220234241@serebryakov.spb.ru> To: freebsd-net@FreeBSD.org In-Reply-To: <12810339411.20101220205327@serebryakov.spb.ru> References: <12810339411.20101220205327@serebryakov.spb.ru> MIME-Version: 1.0 Content-Type: text/plain; charset=windows-1251 Content-Transfer-Encoding: quoted-printable Cc: freebsd-net@FreeBSD.org Subject: Re: E4500 spend one core to saturate 1Gbit/s link with TCP -- is it nornal? 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: Mon, 20 Dec 2010 20:42:49 -0000 Hello, Freebsd-net. You wrote 20 =E4=E5=EA=E0=E1=F0=FF 2010 =E3., 20:53:27: > I have file server with Core2Duo E4500 CPU and Intel gigabit adapter > (82566DM). > FreeBSD 8-STABLE (8.2-PRERELEASE #1: Fri Dec 3 2010) and > samba35-3.5.6 are installed. > This server could serve about 110MiB/s to one client over CIFS, but > one core of CPU is completely busy in such situation: about 60% of CPU > is occuped by "ketnel" and 35-40% occuped bu "smbd" accroding to "top > -S" ouput. > Is it normal, that 2.2GHz core is needed to saturate 1Gib link with > only one client (and one TCP connction), or I have something > misconfigured? Ok, pressing "H" in "top" shows, that all this CPU goes to "{em0 tasq}" thread. And, yes, I have all rxd/trd and other tunables set to recommended high values. --=20 // Black Lion AKA Lev Serebryakov