Next Generation Internet: A work in progress
(IDG) -- The Next Generation Internet (NGI) initiative is picking up steam, fresh from the success of the NGI Forum meeting a few weeks ago.
An inquisitive group of 50 users, vendors and service providers gathered late last month at the University of Virginia's Center for Innovative Technology in Herndon, Va., to explore issues such as IP multicasting, policy-based networking and service-level agreements (SLA). What struck me as odd, though, was the sentiment among some participants that the NGI already exists.
Keynote speaker Richard Mandelbaum, chairman and CEO of Applied Theory, spoke about network intelligence and addressed the issues surrounding the particular layer at which those smarts should reside. Interestingly, his company's Web site touts Applied Theory as "The Next Generation Internet Company."
These remarks about NGI are in contrast with the overriding theme of the day: While we're making great strides toward a unified implementation of the NGI, we aren't there yet. The NGI is a future phenomenon, not quite near the level of maturity required to support ubiquitous user traffic on a national, let alone international, basis.
In fact, the NGI Forum meeting was chock-full of examples that underscored the strides that have been made with NGI, as well as the steps still left to take.
Attendees debated the merits of IP multicasting and where it fits into plans for the NGI. They also questioned whether implementing IP multicasting at the lower layers of a multivendor switch/router network in the NGI is the way to go, or if the problem can be better solved with a mesh of Web-caching proxy servers running over a less intelligent network.
A panel of quality-of-service (QoS) experts produced some mildly worded challenges. However, in the end, the panel's consensus surprised no one: If your applications have tight QoS requirements on throughput, latency, etc., then deploy ATM. But you can bet the IP community will solve such QoS-related issues in a pure IP environment in the near future.
Experts also discussed SLAs from the perspective of how they must evolve -- they need to become more tightly aligned with customers' business and real-time applications as opposed to the coarse, averaged performance metrics common in many of today's reports.
All the questions raised and the issues debated left one point markedly clear: We have a ways to go. Indeed, there may be pockets of technologies available today that will play a role in the eventual NGI amalgamation, but, in and of themselves, they do not form the NGI.
Through continued discussions among users, vendors and service providers, NGI will eventually take shape.
Anderson is chief operating officer of The Tolly Group and serves as Secretary/Treasurer of the NGI Forum.
Building the next Internet
RELATED IDG.net STORIES:
Next Generation Internet Forum takes the application view of QoS
The Tolly Group
|Back to the top||
© 2001 Cable News Network. All Rights Reserved.|
Terms under which this service is provided to you.
Read our privacy guidelines.