The Routing Over Large Clouds Mailing List Archive by date

Cell Relay Retreat>List Archive>month:1996-May> msg00185



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

My personal take on cell switching routers

  • From: Masataka Ohta <mohta@necom830.hpcl.titech.ac.jp>
  • Date: Fri, 31 May 96 11:32:39 JST
  • Cc: ion@nexen.com

Rajeev;

> One important aspect of CSRs, which I haven't seen addressed, is that
> the switched cells flow *through the switch/router box*. This may not be
> the most optimum routing of the cells between the communicating
> end-points.

With CSRs, hosts in different subnets can have a direct ATM connection.

You don't have to make your subnet logical to put two physically
distant hosts in a sigle subnet.

So, unless you intentionally want to make your routing inefficient,
CSRs are put on physical border of contegeous IP subnet and the
path over CSRs are the optimal.

And, if you intentionally want to make your routing inefficient,
then, you can get what you want.

							Masataka Ohta