From owner-freebsd-hackers Sun Mar 24 15:43:44 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id PAA03492 for hackers-outgoing; Sun, 24 Mar 1996 15:43:44 -0800 (PST) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id PAA03486 for ; Sun, 24 Mar 1996 15:43:41 -0800 (PST) Received: from msmith@localhost by genesis.atrad.adelaide.edu.au (8.6.12/8.6.9) id KAA26045; Mon, 25 Mar 1996 10:30:47 +1030 From: Michael Smith Message-Id: <199603250000.KAA26045@genesis.atrad.adelaide.edu.au> Subject: Re: Re(2): Changing Ethernet frame size to 576 bytes? To: rkw@dataplex.net (Richard Wackerbarth) Date: Mon, 25 Mar 1996 10:30:47 +1030 (CST) Cc: taob@io.org, freebsd-hackers@FreeBSD.ORG, msmith@atrad.adelaide.edu.au In-Reply-To: from "Richard Wackerbarth" at Mar 24, 96 07:19:50 am MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Richard Wackerbarth stands accused of saying: > > >In other words, if your link is congested and is losing 20% of the packets, > then those losses make the other two fragments useless too, giving you an > 'effective' loss rate of 60%. > > > > This is bogus arithmetic; lossage is a normally a point event and results in > the loss of one unit datagram around the point, regardless of its size. > > What is bogus about his arithmetic? If the losses are infrequent and not > highly correlated, each loss causes a retransmission of the entire large > packet. This causes the number of bytes retransmitted, and therefore the > effective loss rate, to be multiplied by the fragmentation ratio. There are a couple of major bogosities in his argument : - The assumption that the fragment size is actually the normal MTU (which as has been previously described is not generally the case) - The assumption that the retransmission time for a packet is largely dependent on its size. - The assumption that a 20% hit on /3 fragmented packets will equate to a 60% loss of whole packets. This could only happen if the other two fragments of a packet became magically 'unloosable' after the first was lost. Imagine four barrels, red, green, blue, yellow, lined up. (Colour optional 8). Take a shot at them. Consider a holed barrel a 'lost packet'. Now take twelve smaller barrels, three of each of the four colours. Line them up in random order. Take a shot at them. Consider a holed barrel a 'lost packet' for the barrel's colour. Now take a couple more shots at each setup; think about it a bit 8) -- ]] Mike Smith, Software Engineer msmith@atrad.adelaide.edu.au [[ ]] Genesis Software genesis@atrad.adelaide.edu.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control (ph/fax) +61-8-267-3039 [[ ]] Collector of old Unix hardware. "Where are your PEZ?" The Tick [[