Re: OT: billing multicast makes it a non-starter Re: Clarification on Multicast

From: Radia Perlman - Boston Center for Networking (Radia.Perlman@Sun.COM)
Date: Tue Mar 26 2002 - 13:39:26 EST


What does "OT" at the beginning of the subject line mean?

And to argue a bit with Ran:
>>Part of the operational challenge of IP Multicast is that,
>>at present, as deployed, there is no IGP/EGP split.

Actually, I don't even believe an IGP/EGP split is necessarily the
right thing for unicast routing. (Assuming the phrase "IGP/EGP split"
means "two different routing protocols". If it instead means, levels
of hierarchy, then yes, it seems that for unicast routing, that
seems to help it scale.) But for multicast, I do not believe an IGP/EGP
split helps.

I think the problem with multicast was false starts and complexity
trying to implement somethign with a globally unique, topologically
independent 28-bit address.
I think if the model were explicit root addressing (i.e., (root,address) is
the name of the group)
which is the "single source" model or can be applied to
multisender apps using bidirectional trees rooted at "root"),
then multicast would have been sufficiently straightforward that
it could have been deployed a long time ago.

I don't think an IGP/EGP split helps multicast, since a multicast group
has the potential of having members all over the place.

Radia



This archive was generated by hypermail 2b29 : Mon Aug 04 2003 - 04:10:04 EDT