>>>>> On Tue, 26 Mar 2002 13:39:26 -0500 (EST), Radia Perlman - Boston Center for Networking <Radia.Perlman@Sun.COM> said:
Radia> Actually, I don't even believe an IGP/EGP split is necessarily the
Radia> right thing for unicast routing. (Assuming the phrase "IGP/EGP split"
Radia> means "two different routing protocols". If it instead means, levels
Radia> of hierarchy, then yes, it seems that for unicast routing, that
Radia> seems to help it scale.) But for multicast, I do not believe an IGP/EGP
Radia> split helps.
The story goes that when Lenny Bruce returned, disillusioned, from the
Soviet Union, he reported "it's like the whole country is run by the phone
company."
If there were one phone company in charge of "the Network" then maybe all
routing problems would be shortest (or other constraint) path and scaling
problems. As the world is, the Internet depends on policy decisions among
peers - peers who would without policy refuse to interconnect.
Hierarchy addresses scaling issues; policy at administrative boundaries
addresses the conditions for getting players into the game.
If in talking of an IGP/EGP split folks mean that different path selection
algorithms need to run "inside" and "outside", then I agree that this
shouldn't be axiomatic for IRTF. I think, however, what is meant is that
new routing protocols need to support different policy domains, under
control of autonomous peers. That fits better with the argument that an
"IGP/EGP" split has been an enabling factor in the growth of a unicast
Internet and that the lack of such a split has been an inhibiting factor in
the growth of a multicast Internet.
- Tom Barron
This archive was generated by hypermail 2b29 : Mon Aug 04 2003 - 04:10:04 EDT