All things related to multicast

["Gurcharan S. Khanna" <>]

hi all,

i think a question i have is somewhat related to bill's point here (i hope 
so).
i use the access grid between RIT where there are lots of video streams in
the venue and AUK in Kosovo where they don't have enough bandwidth to
sustain all the streams all the time.

i would like be able to have the client in Kosovo ask for only one of the 
streams
in the venue at a time so it doesn't get flooded. is this possible? with SSM?
with v6 multicast? (are these two related?)

if so, what parts of the network path/devices need to be enabled to do this?

thanks for your suggestions/help,

-gurcharan

On 2/28/11 5:12 PM, Bill Owens wrote:
> On Thu, Feb 24, 2011 at 10:57:30AM -0500, Garry Peirce wrote:
> > Using the frequency of messages on this list and working SAP entries as a
> > rudimentary gauge, the level of interdomain v4 multicast work/content seems
> > low (not to mention v6 which is not yet possible for us).
> I think that the level of interdomain multicast is low, and not likely to 
> increase. So I would like to talk about v6 multicast, since I've grown jaded 
> and find ordinary v4 multicast to be quite boring ;)
>
> Actually, it's because I have been worrying for some time about how the 
> transition to v6 multicast will affect the current users, though they may be 
> few in number. Specifically because we will lose interdomain any-source 
> multicast, which means that the original vic/vat model of sources being able 
> to come and go as they please simply doesn't exist any more.
>
> This means no more "TV guide" function with SAP/SDP, which won't be much 
> missed since the list has long been clogged with advertisements from admin-scoped 
> groups coming from misconfigured sources. Unfortunately it also affects the folks at 
> the Access Grid, based on what I can find out about how their software is set up.
>
> It might be possible to rebuild the AG model so that it could function with 
> embedded-RP, if there were a 'host' for each session with a willingness to 
> operate an RP available to all the participants. That's probably the path of 
> least resistance, though it still doesn't look like an easy job.
>
> The $64 question, then: when will this become necessary? I suspect that the 
> pressure will arise in the APNIC region since they'll be the first to run out of 
> v4, and existing networks will eventually feel the need to go behind NAT in order 
> to keep growing (assuming that they run dual-stack, v4 NAT and v6 native). I 
> don't know when that will happen, though; perhaps we have a few years yet before 
> any of the R&E institutions run out of their existing v4 allocations.
>
> Am I missing anything about the AG model, or any feature of v6 multicast that 
> would allow it to continue working as it has? And are there any other 
> widely-used interdomain applications that also rely on the traditional ASM 
> model?
>
> Bill.

--
-------------------------
Gurcharan S. Khanna, Ph.D.
Director of Research Computing
Office of the Vice President for Research
Assistant Research Professor, GCCIS Ph.D. Program
Director, Interactive Collaboration Laboratory
---
Rochester Institute of Technology
One Memorial Drive
Rochester, New York 14623-5603
Phone: 585-475-7504  ~  Cell: 585-451-8370
Email: 

Web: http://rc.rit.edu
Google Map: http://rc.rit.edu/directions.html