From owner-freebsd-advocacy Tue Jul 13 18:43:23 1999 Delivered-To: freebsd-advocacy@freebsd.org Received: from smtp03.primenet.com (smtp03.primenet.com [206.165.6.133]) by hub.freebsd.org (Postfix) with ESMTP id 4B9A715145 for ; Tue, 13 Jul 1999 18:43:20 -0700 (PDT) (envelope-from tlambert@usr02.primenet.com) Received: (from daemon@localhost) by smtp03.primenet.com (8.8.8/8.8.8) id SAA17430; Tue, 13 Jul 1999 18:42:43 -0700 (MST) Received: from usr02.primenet.com(206.165.6.202) via SMTP by smtp03.primenet.com, id smtpd017411; Tue Jul 13 18:42:40 1999 Received: (from tlambert@localhost) by usr02.primenet.com (8.8.5/8.8.5) id SAA22838; Tue, 13 Jul 1999 18:42:38 -0700 (MST) From: Terry Lambert Message-Id: <199907140142.SAA22838@usr02.primenet.com> Subject: Re: Benchmarking web apps on Apache To: seth@freebie.dp.ny.frb.org (Seth) Date: Wed, 14 Jul 1999 01:42:37 +0000 (GMT) Cc: dwilde1@thuntek.net, billf@chc-chimes.com, freebsd-advocacy@FreeBSD.ORG In-Reply-To: from "Seth" at Jul 8, 99 04:31:40 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-advocacy@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > Hold up a sec. FreeBSD did NOT perform as well. Check the stats again. > The only things FreeBSD beat the other OS in was serving STATIC pages (and > mod_perl handler stuff). The "crucial" tests (dynamic content via cgi's) > showed the other OS to edge out our beloved FreeBSD. Pre-forking would probably make a difference. It might be a pain to make "ps" et. al. ignore the preforked process(es), though... My assumption about the three runs was not what others have come up with (run it until the numbers say what you want), but was, I think, to ensure that everything that was going to be in cache was in cache, as it would be if the server were under heavy traffic to the listed pages. I can think of three ways of speeding up Apache service of static content (one of the places FreeBSD and Linux lost out to IIS on NT in the beake-offs). Similar changes could make a difference to Samba as well (e.g. lazy closes, prebinding of generated data after response but before idle, etc.) Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message