From owner-freebsd-isp Thu Nov 14 14:54:50 1996 Return-Path: owner-isp Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id OAA14223 for isp-outgoing; Thu, 14 Nov 1996 14:54:50 -0800 (PST) Received: from red.jnx.com (red.jnx.com [208.197.169.254]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id OAA14184 for ; Thu, 14 Nov 1996 14:54:34 -0800 (PST) Received: from chimp.jnx.com (chimp.jnx.com [208.197.169.246]) by red.jnx.com (8.7.6/8.7.3) with ESMTP id OAA22572; Thu, 14 Nov 1996 14:50:51 -0800 (PST) Received: (from tli@localhost) by chimp.jnx.com (8.7.6/8.7.3) id OAA04577; Thu, 14 Nov 1996 14:50:49 -0800 (PST) Date: Thu, 14 Nov 1996 14:50:49 -0800 (PST) Message-Id: <199611142250.OAA04577@chimp.jnx.com> From: Tony Li To: stefan@exis.net CC: jdd@vbc.net, jhay@mikom.csir.co.za, chad@gaianet.net, isp@FreeBSD.org In-reply-to: (message from Stefan Molnar on Thu, 14 Nov 1996 17:59:11 -0500 (EST)) Subject: Re: Decision in Router Purchase Sender: owner-isp@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk That was one idea a few months ago, but we got some MegaT ds1 mux units for better threwput, since the cisco can not handle the load ballance of the 2 Ts that are on it. The cisco does load sharing, not load balancing. The MUX (of course), does MUXing and can increase your effective bandwidth. Whether it's cost-effective depends on your personal situation. In general, a cisco will give you about 80% of the bandwidth of two T1's through load sharing. This gets worse if the number of entries in your fast switching cache is low. But from everything that I have seen, from the o'really books, to the lame teachings of my cne lessions, and people from cisco and andersen consulting say a t1 run at 1.54 (after all the overhead). If it was that high, 3.0 then something is wrong in IEEE land. You are correct in that T1 is 1.54Mbps (_before_ overhead), however it's full duplex. Thus, the _aggregate_ bandwidth of the link is ~3Mbps. Tony