*BSD News Article 77508


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!news.rmit.EDU.AU!lynx.aba.net.au!munnari.OZ.AU!news.mel.connect.com.au!news.mira.net.au!vic.news.telstra.net!act.news.telstra.net!psgrain!charnel.ecst.csuchico.edu!newshub.csu.net!newshub.sdsu.edu!newsfeeder.sdsu.edu!news.sgi.com!news.msfc.nasa.gov!newsfeed.internetmci.com!solaris.cc.vt.edu!server.cs.vt.edu!ceharris
From: ceharris@mal.com (Carl Harris)
Newsgroups: comp.unix.bsd.freebsd.misc
Subject: Re: 100BaseT tuning considerations?
Date: 2 Sep 1996 23:31:21 GMT
Organization: Maladjusted Communications
Lines: 23
Message-ID: <50fqo9$fkp@server.cs.vt.edu>
References: <4vin1s$4um@server.cs.vt.edu> <4vv8aj$3sh@ardbeg.islay.sub.org>
Reply-To: ceharris@mal.com
NNTP-Posting-Host: mal9000.mal.com
X-Newsreader: TIN [version 1.2 PL2]

Patrick M. Hausen (pmh@ardbeg.islay.sub.org) wrote:
: What's wrong with that? Have you ever seen 2 high performance Pentium
: servers with Intel Etherexpress Pro/100 running that "industry standard"
: Unix called SCO Open Server 5.0?
: 1.1 MByte per second - you are getting almost five times as much ;-).
: Seriously - 100MBit/s is the available bandwith, not necessarily your
: peer-to-peer performance. As other posters stated, watch your CPU usage
: and experiment with the window size.

I'm mostly interested in determining what factors are limiting the use
of the available bandwidth, not so much in actually using that
bandwidth.   It's precisely the issue that peer-to-peer performance is
not what one would expect from a reasonably high-performance kernel on 
high-performance machines that interests me.

I don't really care how badly SCO performs -- it's beyond hope or
salvation.

--
Carl Harris
Systems Engineer
CNS Research and Planning, Virginia Tech
ceharris@vt.edu