Re: [RRG] December meeting: call for contributions

AFAIK 151 research projects in the category "the network of the future" have qualified for being funded by the EU. So I am curious about the many contributions to be presented to the RRG soon. Unfortunately our research project was rated to be POOR and was therefore rejected. The poor objectives were:

- developing a hierarchical routing protocol which fits for the combined intra- and interdomain network such that it will scale (1-2 K instead of 200-300K FIB entries) for good and that UPDATE-churn is avoided (the more distant the lesser the churn)

- providing multipath forwarding along shortest paths, along detouring but loop-free paths, along loop-endangered paths but providing support to ensure loop-less forwarding.

- real-time traffic balancing based on a p2mp/mp2p message ping-pong between an (overloaded) transit node and precisely all nodes of its specific "ingress sector" wrt to traffic towards a given destination node

- shared network maintenance by a set of arbitrarily spread "controler nodes" based on the fact that if each network node knows the entire topology as well as these controler nodes, any node may determine for every node and every link by which controler node it is to be controled.

- Multihoming in the light of the developed extremely scalable hierarchical routing protocol (see above).

This hierarchical routing protocol wouldn't be prefix-centric any more. Hence IPv6 (remember: factor 4 to make the current situation worse) wouldn't be a problem, nor a longer life of IPv4 as even single IP addresses rather /8 blocks might be taken from the IANA-reserve pool.

I also referred to the so-called Rhekter's law and added that the right choice should be made and that the terrible mistake as to make the wrong choice a second time should be avoided. I am afraid that CONS (etc.) is about to make this terrible mistake. It is a huge machinery for distributing/retrieving routing relevant information which we could get for free - not in the same way, but in a much better i.e. scalable way, which is not afraid of mobile applications.

IMHO: Yakov's remark was not upgraded to a law, but downgraded to a bon mot. In some way, it was never taken seriously - unfortunately.

Heiner

In einer eMail vom 06.09.2007 16:16:04 Westeuropäische Normalzeit schreibt jnc@mercury.lcs.mit.edu:

When it has tried to do just these things, it has succeeded very well. When
it tries to make a pick, it often falls on its face.

Noel