The IP Over NBMA (ION) Archive

[Masataka Ohta <mohta@necom830.hpcl.titech.ac.jp>]

>  >                Multicast Inscalability over Large Cloud

> now, for something to simply "not scale", it has to be exponential

OK. You think both RIP and single subnet ATMARP/MARS scales and you
don't need CIDR nor NHRP. That's fine.

> as a comp sci exponent, i'd like to ask for a proper complexity
> analysis here

I'm afraid a real comp sci exponent won't say:

	"for something to simply "not scale", it has to be exponential".

Anyway, you should read the section: "2. Assumptions".

To be very coarse both NHRP and ATMARP have the complexity of
O(# of hosts in a cloud). A little more detailed analysis shows
that the constant facters are mostly the same.

No analysis on absolute values necessary.

Instead, it is proven that NHRP is only as scalable as ATMARP and MARS.

> network centric:
> a) control messages

> b) memory/state

> c) latency for traffic

> d) throughput,

> and so on.....

May be, you can prove other inscalability of NHRP.

But, see the "abstract" section.

It's not my point to be involved in detailed analysis of NHRP.

One generic proof of inscalability is more than enough.

> and TV doesn't have really low latency requirements, so suboptimal
> placement of the RP or core is not a problem - virtual cores within
> the cloud can be placed easily.....to assign the 'center' to a
> MCS.....or a hierarchical approach such as HPIM or CGBT over a
> hierarchy of MCSs could be used

So, your point is that MARS as is scales, which has no effect on
my proof.

							Masataka Ohta