From owner-freebsd-hackers Thu Feb 4 23:35:24 1999 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA09661 for freebsd-hackers-outgoing; Thu, 4 Feb 1999 23:35:24 -0800 (PST) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from vaio.ispra.webweaving.org ([157.150.122.87]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA09620 for ; Thu, 4 Feb 1999 23:35:02 -0800 (PST) (envelope-from dirkx@webweaving.org) Received: from webweaving.org (localhost [127.0.0.1]) by vaio.ispra.webweaving.org (8.8.8/8.8.8) with ESMTP id IAA02216; Fri, 5 Feb 1999 08:11:54 +0100 (CET) (envelope-from dirkx@webweaving.org) X-MX-Masquarade: Passed MX vaio.ispra.webweaving.org at Vaio / WebWeaving X-No-Spam: Neither the originator(s) address(es) nor the Receipient(s) addresses are to be used for unsolicited commercial email (spam) as a per message fee is incurred for both inbound and outbound traffic Message-ID: <36BA99B9.851BB61D@webweaving.org> Date: Fri, 05 Feb 1999 08:11:53 +0100 From: "Dirk-Willem van Gulik (vaio)" Organization: Web Weaving Consultancy X-Mailer: Mozilla 4.5 [en] (X11; I; FreeBSD 2.2.6-RELEASE i386) X-Accept-Language: zh, nl, en MIME-Version: 1.0 To: Matthew Dillon , freebsd-hackers@FreeBSD.ORG Subject: Re: Irratic Curve References: <36B97B38.74FF0B68@webweaving.org> <199902050005.QAA91618@apollo.backplane.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Matthew Dillon wrote: > > :... > :http://www.scl.ameslab.gov/Projects/Gigabit/performance/prelim.html > : > :Now could any one explain to me WHY freebsd appears so unpredicatable ? > :i.e. not a nice S-curve ? Is that the way of measuring ? Some other > :artifact, or real ? I think it is real, as I get the same sort of > :holes in my graphs for the transaction server. > : > :Any chances on an expose.... > : > :Dw > > If you want a transactional connection over TCP to go fast, you generally > have to turn off nagle. > > { > int one = 1; > setsockopt(fd, IPPROTO_TCP, TCP_NODELAY, &one, sizeof(one)); > } > Actually, I am doing that, as well as various other kernel and buffer mods. But the issue I wanted to bring up was that on Linux and Solarus, you get nice n.log(n) and S shaped curves for rtt/size or speed/size etc. For freebsd that does not seem the case. Now this does not meen that FreeBSD is 'worse'. But I am curious to the background/engineering of this. Why is it not a clear S shaped curve, perhaps with one or two jumps when you cross something like an MTU or ethernet-frame size. Dw To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message