From owner-freebsd-questions Thu Sep 19 12:25:56 2002 Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 14B0B37B401; Thu, 19 Sep 2002 12:25:55 -0700 (PDT) Received: from smtpout.mac.com (smtpout.mac.com [204.179.120.86]) by mx1.FreeBSD.org (Postfix) with ESMTP id 96F0A43E4A; Thu, 19 Sep 2002 12:25:54 -0700 (PDT) (envelope-from cswiger@mac.com) Received: from smtp-relay03.mac.com (smtp-relay03-en1 [10.13.10.222]) by smtpout.mac.com (Xserve/MantshX 2.0) with ESMTP id g8JJPsWI015344; Thu, 19 Sep 2002 12:25:54 -0700 (PDT) Received: from asmtp01.mac.com (asmtp01-qfe3 [10.13.10.65]) by smtp-relay03.mac.com (8.12.1/8.12.1/1.0) with ESMTP id g8JJPrKN024303; Thu, 19 Sep 2002 12:25:54 -0700 (PDT) Received: from bust ([12.38.161.88]) by asmtp01.mac.com (Netscape Messaging Server 4.15) with ESMTP id H2P9Z500.2HP; Thu, 19 Sep 2002 12:25:53 -0700 Date: Thu, 19 Sep 2002 15:25:52 -0400 Subject: Re: WCCP performance questions Content-Type: text/plain; charset=US-ASCII; format=flowed Mime-Version: 1.0 (Apple Message framework v482) From: Chuck Swiger To: , Content-Transfer-Encoding: 7bit In-Reply-To: Message-Id: <9C5B6486-CC05-11D6-801D-000A27D85A7E@mac.com> X-Mailer: Apple Mail (2.482) Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Tuesday, September 17, 2002, at 04:51 AM, Girnet Vladimir wrote: > We are about to build some new WCCP servers. These servers will be based > on FreeBSD 4.6.2 with Squid 2.4.7. > > The hardware will be: 2 x PIII 1.3GHz, 1GB RAM, 100GB HDD. Servers must > serve about 200 requests per second, from about 1000 users. > > I have a question: What performance tuning must be made, to achive the > best results from these servers? Read "man tuning", of course. 200 requests per second is a significant load-- around 2 MB/s I/O bandwidth, given an average WWW request size of ~10K, only remember that you have to read the requested data (at least the first time, until it's cached) and forward the response to the client, so you're both sending and receiving the bytes some of the time. You want to either have a fast SCSI disk system, or multiple spindles, or preferably both. Four 30 GB disks would perform a lot better than one 100 GB disk. With the way squid distributes cache directories, you wouldn't even need to RAID-0 (stripe) them, although that wouldn't be a bad thought either. You could get by with a single 100TX NIC per machine, but having two NICs would be better, since you can easily divide the network traffic up. -Chuck Chuck Swiger | chuck@codefab.com | All your packets are belong to us. -------------+-------------------+----------------------------------- "The human race's favorite method for being in control of the facts is to ignore them." -Celia Green To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message