From owner-freebsd-hackers Fri Nov 30 16:41:37 2001 Delivered-To: freebsd-hackers@freebsd.org Received: from ussenterprise.ufp.org (ussenterprise.ufp.org [208.185.30.210]) by hub.freebsd.org (Postfix) with ESMTP id B0AD737B405 for ; Fri, 30 Nov 2001 16:41:34 -0800 (PST) Received: (from bicknell@localhost) by ussenterprise.ufp.org (8.11.1/8.11.1) id fB10fQ701222; Fri, 30 Nov 2001 19:41:26 -0500 (EST) (envelope-from bicknell) Date: Fri, 30 Nov 2001 19:41:26 -0500 From: Leo Bicknell To: Matthew Dillon Cc: Mike Silbersack , freebsd-hackers@FreeBSD.ORG Subject: Re: TCP Performance Graphs Message-ID: <20011130194126.A969@ussenterprise.ufp.org> Mail-Followup-To: Matthew Dillon , Mike Silbersack , freebsd-hackers@FreeBSD.ORG References: <200111302059.fAUKxrI19553@apollo.backplane.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <200111302059.fAUKxrI19553@apollo.backplane.com>; from dillon@apollo.backplane.com on Fri, Nov 30, 2001 at 12:59:53PM -0800 Organization: United Federation of Planets Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Fri, Nov 30, 2001 at 12:59:53PM -0800, Matthew Dillon wrote: > The transmit side requires more thought. I did write that patch, and > it does work, but it's too messy for my tastes. I would personally > much rather rewrite it to (A) fix the RTT stored in the route tables > and (B) adjust the transmit window based on that, which is a much less > sophisticated patch (and less messy), but ought to work quite well > in regards to transmit buffer management. I think I tried this patch, and found some problems with it. As I recall the problems were with extremely high bandwidth connections (eg, I have two machines that can move 100Mbps FDX across country (70ms latency), and when I tried the patch with that case performance was "bad", in the sense that I got like 20Mbps, rather than 100, like it should have allowed. I believe I theorized at the time that the calculation code had a term that ended up zero, or infinity or something with values that large. Sadly, at this moment I don't have test boxes on that path, but I do now have test boxes behind an otherwise empty T1 that I can do interesting things with (eg, WRED on or off, short queues, long queues, artificial limits like CAR). I have a little time to do testing (emphasis on little), if one or two people would like to do more I can make the resources available, contact me privately. -- Leo Bicknell - bicknell@ufp.org - CCIE 3440 PGP keys at http://www.ufp.org/~bicknell/ Read TMBG List - tmbg-list-request@tmbg.org, www.tmbg.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message