All things related to multicast

["Kevin C. Almeroth" <>]

Personally, a good starting place for v6 multicast is SSM.  Lots of
people will come down on either side of arguing whether "only" should
come after SSM.

At this point, I think vendors can put together a roadmap for v6 SSM.
In some cases, several of the router vendors already support everything
a core router needs.

The stumper is again switch support for snooping MLDv2 and of course
host support.

-Kevin

On Wed, 11 Aug 2004, Brent Sweeny wrote:

>>This April note from Bill (plus some clarifications from Dino) seems
>>to have been the clearest explanation of the nature and status of
>>embedded-RP.
>>
>>current ietf version is:
>>http://www.ietf.org/internet-drafts/draft-ietf-mboned-embeddedrp-07.txt
>>(expires Jan 2005)
>>
>>In today's JET meeting, the folks from DREN expressed some interest
>>in the NREN efforts to move forward with wider deployment of inter-
>>domain v6 multicast and are asking what they should tell the vendors.
>>Is something like this, and urging them to commit (and show on their
>>roadmaps) it?    I'd like to come up with a good statement on why
>>this is good, why it's the standard direction accepted by at least
>>the R&E networking community as much as represented here, and at most
>>the technical multicast and v6 people who understand the issues.
>>(and if it's NOT those things, what is?)
>>      thanks/ Brent
>>
>>On Thu, Apr 01, 2004 at 10:35:18AM -0500, Bill Owens wrote:
>>> On Thu, Apr 01, 2004 at 09:35:00AM -0500, Michael H. Lambert wrote:
>>> > Could someone please summarize embedded-RP to the list so that those of 
>>> > us who haven't been paying any attention to IETF multicast efforts will 
>>> > gain some clew?
>>> 
>>> The idea is straightforward, but ingenious; configure an RP at a special 
>>> address within your IPv6 space, one that can then be encoded into a few 
>>> bits that become part of the group address. It builds on RFC 3306 
>>> "Unicast-Prefix-based IPv6 Multicast Addresses". It lets you generate 
>>> globally unique group addresses, and avoid flooding source information 
>>> (so no more MSDP) or having to discover sources out of band (SSM).
>>> 
>>> The current version is draft-ietf-mboned-embeddedrp-02.txt. There's been 
>>> quite a bit of discussion about it on the IETF mboned list in the last 
>>> couple of weeks, but mostly around how it interacts with bidirectional 
>>> PIM (which I don't fully understand). Looks like it will go to last call 
>>> soon, and start on the standards track.
>>> 
>>> The only implementation I now of is in IOS, you can find the versions 
>>> with Feature Navigator. I have a router that's running it but haven't 
>>> lashed up a test yet. . .
>>> 
>>> Bill.
>>