All things related to multicast

[Kevin Thompson <>]

At 06:06 PM 9/26/99 -0700, Greg Shepherd wrote:

> cisco hosted a multicast engineering workshop for I2 earlier this summer.
> I believe this overview/case-study approach worked very well.
> This format would work well for a WG meeting also.


I thought Cisco's event was very helpful to those who attended


> We could possibly take case-studies submitted in advance to ensure a
> representative sample with various topoligies/technologies.


good idea

kevin


> Greg
>
> On Sat, 25 Sep 1999, Kevin C. Almeroth wrote:
>
> > >>    What good does it do to deliver multicast to a campus when the 
> campus network
> > >>is not engineered to handle delivery of that traffic?  Is anyone 
> doing something
> > >>that is working?  What approaches to this problem exist?  What are 
> the local
> > >>engineering/technical issues?
> >
> > This kind of problem is tough to cover at this kind of member meeting.
> > It is more designed for a Techs Workshop.  Who from your campus has been
> > attending these?  In fact, there have already been a number of events
> > which focus on exactly the questions you are asking.  Personally, I
> > believe we are still a long way away from having good, self-contained
> > answers but that's a near impossibility when just about every campus
> > is different.
> >
> > >>    I am a member of ViDe; some of my fellow committee members are 
> convinced that
> > >>practical delivery of multicast will NEVER work.  I don't happen to 
> agree with
> > >>them, but I don't see anything proposed here that would make them want to
> > >>reconsider.
> >
> > Frankly, if they aren't willing to learn there isn't much anyone can
> > do about it.  The way multicast (or any other service for that matter)
> > typically gets deployed is if someone on the campus makes the effort
> > to learn how to get it running.   Granted, the amount of available
> > information is somewhat sparse, but that problem is improving.  Right
> > now, it takes just a little bit of extra effort.  And a fair number of
> > campuses and gigapops have demonstrated it is possible (see I2 multicast
> > WWW page for maps with lists of connected sites).
> >
> > Bottom line is if you (or an appropriate representative) has technical
> > concerns the "backbone engineers" session would be appropriate.
> > Otherwise, show up at the "neophytes" meeting.  There will be a
> > significant number of people in a similar situation.  A driving question
> > for me personally (as WG chair) is how to provide the kind of information
> > you need in a way that you (or any campus) can use it.
> >
> > BTW, in retrospect, I think "neophyte" is exactly the category you
> > would fit into...  not a novice but also not an expert.  Below, I
> > cut-and-pasted the text from the agenda about what the "neophyte"
> > session is supposed to cover:
> >
> > Multicast: A discussion of how the support of high-speed
> > native multicast by Internet2/NGI backbones is requiring upgrades to
> > campus/GigaPoP networks as well as work on multicast-capable applications.
> >
> > -Kevin
> >
> >