Something better than IPv6
Johna Till Johnson at Network World cleverly compares IPv6 to the U.S. Financial crisis in this article. She believes that IPv6 may well have been an ill conceived solution to the problem and cites John Day's research into a whole new solution to the problem. (click for an introduction to Day's ideas's)
John Day is not the first guy to suggest we build something new from scratch. Forgetting the world of IETF for a moment, Todd Underwood proposed it at a conference a couple of years back, which I referenced in a blog shortly thereafter (Apr 2006) .
One of the points that I made then which I still believe now is that there is going to be a shortage of IP addresses in the future and it may be too late to start working on something new. Personally, I do not have faith that the IETF can work fast enough to create a brand new protocol that will solve the shortage while providing new and interesting functionality. Once they get it implemented, then every router and operating system vendor in the world must update their software to support the new protocol, extensive interoperability tests have to be performed, etc. etc. Then you have to get the entire world to deploy it! This is only something that could be whipped out in 2-3 years if everyone was in agreement that a simple and straightforward stop-gap measure was necessary, and that would be hard since IETF is largely a meritocracy that demands consensus among academics, technologists and vendor interests before it can even hope to move out of Internet-Draft into RFC (standard).
Having said that, Global Crossing would do what is necessary to implement the new protocol on our network. As with our IPv6 deployment, we may not even have to upgrade the core of the network, just the edge platforms which may give us an edge in cost and deployment speed over some of our competition. This is because we are one of the few backbones that switch 100% of our Internet, IP-VPN and VoIP traffic using MPLS in our core.
What should we do? Should we change our direction? Nah, I don’t think so. IPv6 is the only thing out there and we’re ready to support it right now. There is no “gee, maybe we should have done xxx instead" because xxx doesn't exist.
What I will say is this. You got a better idea? Show us and we’ll do it.








