From owner-freebsd-stable Thu Dec 13 5:41:37 2001 Delivered-To: freebsd-stable@freebsd.org Received: from smtp.www-service.de (smtp.www-service.de [212.77.161.16]) by hub.freebsd.org (Postfix) with SMTP id D0BF137B6E6 for ; Thu, 13 Dec 2001 05:29:52 -0800 (PST) Received: (qmail 10693 invoked from network); 13 Dec 2001 13:29:50 -0000 Received: from pd9503389.dip.t-dialin.net (HELO fw.tue.le) (217.80.51.137) by smtp.www-service.de with SMTP; 13 Dec 2001 13:29:50 -0000 Received: from mezcal.tue.le (mezcal.tue.le [192.168.201.20]) by fw.tue.le (8.8.8/8.8.8) with ESMTP id OAA12669; Thu, 13 Dec 2001 14:29:35 +0100 (CET) (envelope-from thz@mezcal.tue.le) Received: (from thz@localhost) by mezcal.tue.le (8.11.6/8.11.6) id fBDDTYK00840; Thu, 13 Dec 2001 14:29:34 +0100 (CET) (envelope-from thz) Date: Thu, 13 Dec 2001 14:29:34 +0100 From: Thomas Zenker To: sthaug@nethelp.no Cc: stable@freebsd.org Subject: Re: TCP stack still hosed? Message-ID: <20011213142934.A771@mezcal.tue.le> Mail-Followup-To: Thomas Zenker , sthaug@nethelp.no, stable@freebsd.org References: <20011213134244.A380@mezcal.tue.le> <32389.1008248631@verdi.nethelp.no> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <32389.1008248631@verdi.nethelp.no>; from sthaug@nethelp.no on Thu, Dec 13, 2001 at 02:03:51PM +0100 Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Thu, Dec 13, 2001 at 02:03:51PM +0100, sthaug@nethelp.no wrote: > > repeated the tests with a normal ethernet card. I had to change the > > test procedure, because of the 100Mbit ethernet sysinstall was masking > > the results. So I did fetch ftp:// of the iso image, about 150MB. > > > > the result is different: > > server 2001/12/13, client 2001/12/12: 6 MBps > > server 2001/11/20, client 2001/12/12: 5.8 MBps > > server 2001/11/20, client 2001/07/17: 7.2 MBps > > > > so here the differences are not that much. > > How close are these hosts? As far as I can remember, the TCP problem > that was corrected related to behavior under packet loss. > > With two hosts running -STABLE connected with a switch I have no problem > filling a 100 Mbps Ethernet pipe: > > mgmtserv2# ttcp -r > ttcp-r: buflen=8192, nbuf=2048, align=16384/+0, port=5001 tcp > ttcp-r: socket > ttcp-r: accept from 195.18.128.227 > ttcp-r: 134217728 bytes in 11.35 real seconds = 11543.38 KB/sec +++ > ttcp-r: 91930 I/O calls, msec/call = 0.13, calls/sec = 8096.18 > ttcp-r: 0.0user 0.4sys 0:11real 4% 15i+206d 244maxrss 0+1pf 91929+2csw > > 11544.62 * 8192 * 1538/1460 = 99.626 Mbps. Close enough :-) > > Steinar Haug, Nethelp consulting, sthaug@nethelp.no Yes, the numbers above are with ftp, so there might be also some loss in performance. The problem I have, as you can see in my first message, is the regression from about 240 Kbyte/sec with two system from before 20 nov. and 15 Kbyte/sec with the stable from today. There might be a problem with lossed packets, as the server has 100Mbps and the client a 10Mbps ethernet. Thomas -- Thomas Zenker c/o Lennartz electronic GmbH Bismarckstrasse 136, D-72072 Tuebingen, Germany Phone: +49-(0)7071-93550 Email: thz@lennartz-electronic.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message