Date: Fri, 17 Apr 1998 19:46:52 -0700 From: Don Wilde <dwilde1@ibm.net> To: rse@engelschall.com Cc: freebsd-advocacy@FreeBSD.ORG, jkh@time.cdrom.com Subject: Re: SPECweb96 challenge Message-ID: <3538141C.B067E439@ibm.net> References: <35363396.CF5289F3@partsnow.com> <19980417123801.A9264@engelschall.com> <35376D98.7676B8AA@ibm.net> <19980417193543.A14608@engelschall.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Ralf S. Engelschall wrote: > > On Fri, Apr 17, 1998, Don Wilde wrote: > > > > > > Novell Web Server Pushes Envelope > > > > > > > > > > Novell, Inc. reached a SPECweb96 benchmark of 1639 running > > > > > intraNetWare 4.11 on a single processor Compaq ProLiant 3000. > > > >[...] > > > > 1) Obtain SPECweb96 test guidelines > > > > 2) Obtain equivalent test platform hardware > > > > 3) Configure FreeBSD for max serving performance > > > > 4) Configure Apache for max serving performance > > > > 5) Run test > > > > 6) Promote results heavily!!!!! > > > > > > > > My best hardware is a PPro180 with 64MB, so I don't have a new enough machine, > > > > but I think this is exactly the kind of thing we FreeBSD advocates need to do to > > > > promote FreeBSD -- and get the IW's and NWW's and Byte's to acknowledge us. If > > > > we've got TBDOSITW, let's _prove_ it. > > > > > > There is only one point you missed: the SPECweb96 benchmark costs $800 :-( > > > > Between apache and freebsd we can't find $800? I'm going to ask Jordan > > if we can approach this from the central POV by putting out an official > > challenge call on the site and on -questions, -newbies and -chat, and > > where we can send the money. Ralf, will you find out the rest of the > > particulars as to what we need to actually perform a rigorous test that > > we can put side-by-side with the payware geeks? > >[...] > > Beside from the fact that I currently don't have time for any such stuff (I'm > totally busy with Apache 1.3b6 and now 1.3b7) I'm not sure if your idea is > really useful. Benchmarks are problematic in _a lot of cases_ and the final > gain for FreeBSD+Apache is not obvious, IMHO. Everyone knows that we are doing > a great work with FreeBSD and a great work with Apache. And everyone knows > that both products work nicely together and form a powerful solution. But if a > SPECweb96 benchmark result really is worth the effort of $800 I'm not sure. > > Instead I really think that a benchmark which costs $$ is a bogus one. It's > just there for making money instead of making good results. So, it would be > perhaps better to promote FreeBSD+Apache via other alternatives. But > nevertheless if someone contributes $800 and the hardware plus time to run > such a SPECweb96 benchmark with FreeBSD+Apache I'm interested in the results, > of course. But I think it is not worth the effort... I saw reference to a WEBstone benchmark on SGI's site. I'll go dig up some info on that, and get back to you. SPEC is a commercially-backed consortium, but it IS a known entity, even if they constantly fight over what is the benchmark. My point is that it is sometimes important to beat them on their own turf. Perhaps another good entry point would be the new InfoWorld Testing Labs, although they do seem to be rather mercenary (!) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3538141C.B067E439>