All things related to multicast

[Leonard Giuliano <>]

On Fri, 5 Sep 2008, Hitoshi Asaeda wrote:

-) > An attempt was made several years ago to deprecate ASM for IPv6 
-) > (draft-giuliano-mboned-v6mcast-framework).  This proposal was not warmly 
-) > received in MBoned (baby-seal-clubbing would have been more popular at 
the 
-) > meeting).
-) > 
-) > Embedded RP has pretty much guaranteed that we will have to live with 
ASM 
-) > in IPv6 for the forseeable future (forever).
-) 
-) We should have kept possibility of ASM communication model in
-) IPv4/IPv6, because ASM is useful for service discovery (e.g. SAP,
-) SLP). Well, ND works with ASM, too. We cannot eliminate ND from IPv6.
-) But such service discovery is assumed or should be used in a LAN or a
-) small and controled domain with a single policy (e.g. in an ISP).
-) There is no use case that encourages to use such service discovery in
-) a large scale interdomain network. And for streaming applications, SSM
-) fits the demand and there is no reason to live with ASM.
-) 
-) We should classify and distinguish the use cases of ASM and SSM.

The problem is that IP services don't like borders.  Inevitably, any 
service that finds a niche on LAN/small domain will find it's way out 
onto the Internet.  This is, in fact, the whole idea behind the Internet- 
what makes the Internet so great.