The IP over ATM Mailing List Archive by date

[jed@llnl.gov (James E. [Jed] Donnelley)]

I raised the concern below on the multicast oriented
mail lists that I am familiar with (mbone@isi.edu and
rem-conf@es.net).  It was pointed out to me that this
list with its focus on IP over ATM might be an appropriate
place to look for feedback also.  I should probably note
that my concern is not specific to IP over ATM, but
to multicast ATM (whether IP multicast or "native"
ATM multicast).  Still, any feedback is appreciated.
__________________________________________________
I would like to raise a concern I have about approaches
and progress on multicast over Asychronous Transfer
Mode (ATM) technology (adapters, switches, etc.).  I will
admit upfront that I do not have complete knowledge of
where progress is being made in this area.  It is an
area that I dabble in but am not actively involved
in.  I could just let my concern go, but decided to try
to clarify it with this message.  I'm ready to be blasted,
but I hope I'm blasted with some pointers to some new
information on multicast over ATM.

Having read the Internet-Draft "Support for Multicast
over UNI3.1 b ased ATM Networks" - draft-ieft-ipatm-07.txt,
I am concerned about the direction that seems to be being
taken.  In that document there are two approaches described
for routing multicast packets (independent of how multicast
groups are set up):

1.  Meshes of point to multipoint Virtual Circuits (VCs), or

2.  routing multicast packets through a multicast server (MCS)
that has a single point to multipoint VC to all the leaf notes
of a multicast group.

Both of these approaches seem quite poor to me.  They are both
serious steps backwards from the sort of mechanisms
that are currently available for IP multicast (either on MBone or
in multicast capable routers).

It seems to me that what is needed is some sort of multicast
virtual circuit.  With such a circuit, nodes could join locally
and have their cells routed (through a table in the ATM switches
much like with unicast or point to multipoint VCs) to the appropriate
next switches (much like as now happens with IP multicast).
Specifically cells would not be routed back to where they came
from (e.g. as happens with the MCS approach).  Also, only one
such structure would exist in any switch for a multicast group
(unlike with the mesh of VCs approach which has one point to
multipoint VC for every leaf node transmitter in a multicast group).

Can anyone give me a pointer to work on ATM multicast beyond
that described in the draft-ieft-ipatm-07.txt?  It would seem that
such work would require a UNI beyond 3.1 with perhaps some
support for direct multicast VSs?  Any pointers and/or suggestions
are welcome.  Thanks!


--Jed   http://www-atp.llnl.gov/atp/jed-signature.html