From owner-freebsd-hackers Tue Jul 6 11:42:33 1999 Delivered-To: freebsd-hackers@freebsd.org Received: from cs.rice.edu (cs.rice.edu [128.42.1.30]) by hub.freebsd.org (Postfix) with ESMTP id 9597F15003 for ; Tue, 6 Jul 1999 11:42:27 -0700 (PDT) (envelope-from aron@cs.rice.edu) Received: (from aron@localhost) by cs.rice.edu (8.9.0/8.9.0) id NAA12538 for freebsd-hackers@freebsd.org; Tue, 6 Jul 1999 13:42:26 -0500 (CDT) From: Mohit Aron Message-Id: <199907061842.NAA12538@cs.rice.edu> Subject: paper on improving Webserver performance To: freebsd-hackers@freebsd.org Date: Tue, 6 Jul 1999 13:42:26 -0500 (CDT) 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-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Hi, I'd like to tell the FreeBSD community about my recent paper that proposes some TCP implementation techniques for improving Webserver performance. I used FreeBSD for my study and I thought perhaps FreeBSD's TCP maintainers might like to incorporate some of the propositions in FreeBSD. The abstract of the paper is attached below and the paper is available from: http://cs-tr.cs.rice.edu/Dienst/UI/2.0/Describe/ncstrl.rice_cs/TR99-335/ - Mohit Aron aron@cs.rice.edu Abstract: This paper studies the performance of BSD-based TCP implementations in Web servers. We find that lack of scalability with respect to high TCP connection rates reduces the throughput of Web servers by up to 25% and imposes a memory overhead of up to 32 MB on the kernel. We also find that insufficient accuracy in TCP's timers results in overly conservative delays for retransmission timeouts, causing poor response time, low network utilization and throughput loss. The paper proposes enhancements to the TCP implementation that eliminate these problems, without requiring changes to the protocol or the API. We also find that conventional benchmark environments do not fully expose certain significant performance aspects of TCP implementations and propose techniques that allow these benchmarks to more accurately predict the performance of real servers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message