On Thu, 15 Sep 2005 19:34:47 -0700 "Hallam-Baker, Phillip" <pbaker@xxxxxxxxxxxx> wrote: > I think that the reasonable question to ask here is 'does MBONE have a > future?' It doesn't even have a present. It's long gone. Native multicast on the public Internet seems to continue to have a future, but only for a continued subset of the global Internet. I've heard some talk of renewed interest in multicast deployment, but I'm skeptical. If you look at some statistics [1] you find: About 450 ASNs (and not just .edu's) representing about 5000 routes seen via MBGP. These routes expand to cover about 125,000,00 million possible addresses (compared to about 1.1 billion addresses covered by a current unicast table). There are about 1,300 (not totally and obviously bogus) SAs. What is interesting is that these numbers haven't moved all that much over the past few years. Multicast deployment has peaked, but one might consider 10% (using address coverage as one measure) to be pretty widely deployed. What seems to have pretty strong and probably growing interest is site local multicast applications such as file/image distribution and Video over IP. > If I was asked to design a new media distribution protocol > from scratch I certainly would not choose MBONE as a model. The > technical and political problems both appear insoluble to me. They are. Even after the MBONE, large-scale multicast is still quite hard. I think this paper captures the general sentiment pretty well and I bet a few other operators who have had similar multicast deployment experience will be nodding their heads with a familiar feeling of shared pain as they read through Section 2: <http://gravity.lbl.gov/grbell/docs/RIPQOS-GRBell.pdf> Even if you could get past the complexity issue, I would contend that you would also need to overcome the security fears real (and they are in the multicast case) or imagined whenever it requires the netops people to have turn 'something on'. Today, it is more likely that things get turned off, rather than turned on and that has real serious implications for how new things get deployed, which I'm sure is no secret to you and most others here. [1] http://www.multicasttech.com/status/ John _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf