Date: Fri, 05 Feb 1999 08:11:53 +0100 From: "Dirk-Willem van Gulik (vaio)" <dirkx@webweaving.org> To: Matthew Dillon <dillon@apollo.backplane.com>, freebsd-hackers@FreeBSD.ORG Subject: Re: Irratic Curve Message-ID: <36BA99B9.851BB61D@webweaving.org> References: <36B97B38.74FF0B68@webweaving.org> <199902050005.QAA91618@apollo.backplane.com>
next in thread | previous in thread | raw e-mail | index | archive | help
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
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?36BA99B9.851BB61D>