A Practical Navigator for the Internet Economy

ISP Certification Yields New Business Model -- Manager of ANX Overseer Explains Procedures by Which Large ISPs Will Launch Auto Industry VPN

Two Year Long Project of AIAG Ready to Use Public Internet as Platform for Mission Critical VPN with Parts Suppliers -- pp. 1 - 6

We interview Bryan Whittle, General Manager, ANX Overseer, about the startup of the Automotive Network eXchange (ANX). The ANX is a VPN run over the public internet on behalf of the Automotive Industry Action Group (AIAG). ANX will require all automotive industry parts suppliers wishing to business with the Big Three to participate. Depending on one's definition of parts supplier, we are talking between three and ten thousand companies. ANX will be a mission critical network service run only by ANX Overseer Certified ISPs.

The Overseer (contract held by Bellcore) is the neutral coordinating entity connecting ISPs with trading partner customers. It will certify the ISPs that provide the service and serve as the collection point for data measuring over 100 different service criteria that must be maintained by those providing the service.

Given the time critical nature of the conduct of globally interconnected business, certified ISP delivery of mission critical services will, we believe, come to dominate business-to-business use of the internet very quickly. We are looking at a new internet business model that, if it can be priced at a reasonable level, is likely to become VERY popular.

Here Bellcore has positioned itself very very well. Why? Because there are other industries that will likely see an advantage in moving dedicated VPNs to the public internet - if such a move can be made reliably and with requisite security. Financial services and banking, other hi-tech manufacturing like the aircraft industry and tele-medicine come to mind. If ANX works for the auto makers, it is very likely that these other industries will want something similar.

ANX certified ISPs must provide eight categories of service. The first class is composed of services that are required to be offered by certified service providers. Among these are connectivity to trading partners. Any trading partner must be able to use the ANX to reach any other trading partner. Inter-operability metrics make up the second category. These mostly concern connectivity requirements among the service providers. For example: between any two trading partners there can be at most two certified service providers.

Performance Metrics is the third category. For example, file transfer delay. The Overseer will take active measurements to ascertain if a file of a certain size gets across the network of a service provider of a certain size within a specified period of time. Reliability - the availability of the service provider network - is the fourth category. The fifth category is business continuity and disaster recovery. This focus is on the service providers demonstrating they have a robust plan for avoiding disasters and in the event of such a disaster ensuring trading partner continuity of service. Security metrics are the sixth category. Given that trading partners will be using IP sec encrypted data, the focus of the security metrics will be on denial of service attacks. Customer care metrics is the seventh category. For example the performance of the help desk. The eighth and final category is trouble handling metrics.

The pilot phase of the program involving a handful of ISPs and 25 to 35 trading partners is to kick off by the end of the year and run for about four months. Release one will last for about a year and be North American in scope. With Release Two, it is anticipated that ANX will expand globally.

Editorial: John Sidgemore - Would be Internet King? pp. 6, 24

We look with dismay at what the successful acquisition of MCI by WorldCom would mean for the Internet. Can anyone imagine Vint Cerf reporting to John Sidgemore, an economics BA and turn around specialist who has little reason to know about the cooperative nature that made the net possible.

Does IPv6 Have a Future? pp. 7 - 12

Discussions on NANOG examine the difficulty of converting to IPv6, which is found by quite a few to be a solution in search of a problem, as well as a protocol with a European OSI bias. The general consensus of the discussion was that NAT boxes are getting such good capabilities as to be the best answer to problems of routing announcements and IP address portability.

Unfortunately the NAT boxes are not transparent to protocols like CuSeeMe and may also derail IP sec and DNSSEC. There was considerable debate as to what can be done about these problems. The debate ranged from the need to replace BGP4 rather than IPv4 - all the way to using NAT boxes to change the routing structure of the net. While Europe seems very interested in pushing ahead with IPv6, the betting was that given the vast size of the IPv4 installed base, IPv6 networks on this side of the Atlantic would be restricted to tunneling through and IPv4 infrastructure for a long while.

Internet Governance Wars: We Find IWG & IPOC Anatgonism to Network Solutions is Only Crisis --Divergent Agendas Endanger "Bounded Chaos" Nurturing Creative Foundation of Internet Technology Development -- Magaziner Poorly Served by Kahin, Burr and Nelson -- pp. 13 - 22

Motivated by an abiding hatred of Network Solutions' "monopoly" the federal Interagency Working Group on DNS (IWG) ironically adopted, at the end of the summer, an attitude that since the ISOC CORE effect in Geneva was the only other game in town, it should have the US government embrace the very folk that the IWG had been set up in fear of.

We have been told that the idea of assistance in the creation of the database software needed by CORE has been tried out in negotiations for changes that would make the CORE Swiss cartel more 'democratic'. Having been warned that Brian Kahin was holding behind the scenes discussions with IBM, AT&T and Oracle on the question of the data base software and that Becky Burr, Kahin's partner at NTIA, was testing the waters for the idea that a competition for software design be held, we submitted a FOIA to OSTP on October 7 seeking all relevant data to the IBM, Oracle and AT&T discussions. On the 27th we were told that more time would be need because of the "substantial interest in the determination of our request." When on Friday November 14 no documents had been delivered, we complained and were told that the reason was because they involved coordination with the FCC since we had mentioned Mike Nelson, an ex OSTP employee. However by the afternoon the story changed to merely that there were too many and they were waiting to send them all.

During the week that began on November 10th, the Working Group was making its own leaks that it would have its findings released - first by the 14 and then by the 17th. We complained in a public appeal to Ira Magaziner on the 12 and on the 13th received a phone call in which Mr. Magaziner assured us, for attribution, that no policy would be made for a number of weeks yet and perhaps even for several months. On the basis of our conversation we believe that: 1. he is the 'boss' of the IWG and will prevent Burr and Kahin from doing an end run around him in their desire to kill the NSI monopoly. 2. That he has a sincere and humble recognition that he simply doesn't understand the matter well enough yet to reach the right decision. We are encouraged to hear that he will be doing heavy duty fact finding of his own.

Coming to understand much better the reasons for Tony Rutkowski's dislike of the ITU, after the Secretary General admitted that he would like to have the ITU play a major role in Internet governance, we believe that it is critical for Magaziner to understand that the Internet's creative vitality comes from Einar Stefferud calls its self-organizing 'bounded chaos'. Just as no one entity can control the global economy, so no single entity can control the Internet. Stef has a profound understanding of these issues and the final third of our article is a compendium of his recent writings on the subject plus a concluding section that he wrote on November 15 in response to our findings about the working group activities. Here he states that Mr. Magziner must realize that there is no crisis; and that no new top level domains should be added to the root servers until all parties have joined into a confederation using IETF processes to develop their own governance plan. He concludes by pointing out that the ultimate recourse lies in the power of thousands of DNS administrators world wide to point to the root machines that offer the best service to the network.