From owner-freebsd-questions  Fri Sep 26 22:28:28 1997
Return-Path: <owner-freebsd-questions>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.7/8.8.7) id WAA27851
          for questions-outgoing; Fri, 26 Sep 1997 22:28:28 -0700 (PDT)
Received: from spoon.beta.com (root@spoon.beta.com [199.165.180.33] (may be forged))
          by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id WAA27845
          for <questions@freebsd.org>; Fri, 26 Sep 1997 22:28:22 -0700 (PDT)
Received: from spoon.beta.com (mcgovern@localhost [127.0.0.1])
	by spoon.beta.com (8.8.5/8.8.5) with ESMTP id BAA12503;
	Sat, 27 Sep 1997 01:28:02 -0400 (EDT)
Message-Id: <199709270528.BAA12503@spoon.beta.com>
To: beef@tht.net
cc: questions@freebsd.org
Subject: RE: serious PPP problem
Date: Sat, 27 Sep 1997 01:28:02 -0400
From: "Brian J. McGovern" <mcgovern@spoon.beta.com>
Sender: owner-freebsd-questions@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

This is a typical ISP flub. I strongly suspect that their routing
is to blame. Chances are that they're real router (ie - not the 
portmaster) has the route for 209.47.145.1 being sent to the first
portmaster. Both portmasters then have their default route set to their
router (what one would consider normal).

Assuming these are Livingston boxes, I have seen problems with RIP packets
not being generated/eaten by these boxes properly.

My guess, when you connect to the second portmaster, were to to "go out
on the net", and traceroute back to your IP, would find that the ISP's
router is sending all your traffic to the first portmaster, even though
you're connected to the second. Since the first one has no clue as to
how to route the data to the second, the bits end up in the bit bucket.

The way I kinda "hacked" around this was to set the default route for each
of the portmasters to the next in the chain, so that packets destined for
"somewhere" would try all of the portmasters before heading back towards
the default router. This, however, is a flawed solution for a number
of reasons. It just gets around the problem.

Have your ISP help you run traceroutes from outside the segment the
portmasters are on. Chances are, you'll find all your packets going
to the first one.
	-Brian