Path: utzoo!attcan!uunet!lll-winken!lll-tis!ames!oliveb!jerry
From: jerry@oliveb.olivetti.com (Jerry Aguirre)
Newsgroups: comp.dcom.lans
Subject: Re: RetixGate 2244 MAC Bridge - (nf)
Message-ID: <25335@oliveb.olivetti.com>
Date: 13 Jul 88 02:32:47 GMT
References: <281@ifistg.UUCP> <3059@Portia.Stanford.EDU>
Reply-To: jerry@oliveb.UUCP (Jerry Aguirre)
Organization: Olivetti ATC; Cupertino, Ca
Lines: 27

In article <3059@Portia.Stanford.EDU> morgan@Jessica.stanford.edu (RL "Bob" Morgan) writes:

>They are coming out with a bridge management package that runs on a PC
>and is ISO-based.  Haven't seen it, but it looks like an interesting
>design, and it should greatly increase the value of the product.  Any
>network box without management hooks is an invitation to headaches.

I have trouble understanding this statement.  Do you think that a
repeater should have nework management hooks?  Isn't a bridge just a
slightly smarter repeater?  (I am assuming an ethernet to ethernet
bridge here.  Not one using slower intermediate media.)

Assuming the bridge is doing only ethernet address filtering and doesn't
do multicast or other user specified filtering, do you need to manage it
any more than you would manage a repeater?  Granted without special
filtering you have a large network that is vulnerable to many problems.
But a bridge is not a router and even the best of management hooks is
not going to fix that.  The bridge can isolate traffic and provide some
security from monitoring (but not spoofing).

So, if one views a bridge as an inteligent repeater then what management
tools are needed?  The only one I can think of is statistics and that
could probably be better generated from other sources.  I mean, do you
really trust the bridge to tell you about its own performance?

Given the high cost of repeaters, why would someone want to use one
instead of one of the Retix bridges?  (An honest question, I would like
to see the tradeoffs discused.)