Web Perforamnce Versus MTU

[Marc Slemko <ma...@znep.com>]

>Path: scanner.worldgate.com!news.he.net!Supernews60!supernews.com!peerfeed.ncal.verio.net!news1.best.com!sdd.hp.com!col.hp.com!news.cup.hp.com!raj
>From: foo@bar.baz (Rick Jones)
>Newsgroups: comp.benchmarks,comp.protocols.tcp-ip
>Subject: Web Perforamnce Versus MTU
>Date: 23 Feb 1998 22:47:16 GMT
>Organization: http://www.cup.hp.com/netperf/NetperfPage.html
>Lines: 16
>Message-ID: <6c...@ocean.cup.hp.com>
>NNTP-Posting-Host: loiter.cup.hp.com
>Summary: Web Server performance is directly related to link MTU
>Keywords: web server,performance,MTU
>X-Newsreader: TIN [version 1.2 PL2.13]
>Xref: scanner.worldgate.com comp.benchmarks:18905      comp.protocols.tcp-ip:30633     

There is a new "whitepaper" available from
ftp://ftp.cup.hp.com/dist/networking/briefs/ that shows one cut at the
relationship between web server performance and link MTU (Maximum
Transmission Unit - a.k.a. "packet size"). Select your choice of
WebMTU.html, WebMTU.ps or WebMTU.pdf.

The whitepaper may be of some use to those who are (perhaps making
unwitting mistakes in) comparing web server benchmarks between systems
using different link technologies and this different link MTU's.

happy benchmarking, as always, you mileage may vary :)

rick jones
--
these opinions are mine, all mine; HP might not want them anyway... :)
feel free to email, or post, but please do not do both...