A Practical Navigator for the Internet Economy

We have published the August issue of the COOK Report on Internet-> NREN today. Publication is earlier than planned to get the CIX story out to our hardcopy readers. We have also updated the CIX story since our publication last weekend. What follows we have not previously published to the Net -- a summary of the entire story with an emphasis on the context in which the CIX decisions were taken. The remainder of the August issue contains the second installment on the Russian Internet and and exclusive interview with Dr Alexander Goldfarb of the International Science Foundation.

CIX Board Enforces Routing -- Path Routing Filters to Go Up Nov. 1, 1994

Faced with the opening of the Network Access Points on August 15 - what some see as federally inspired competition to its interconnect services - the CIX finally decided to meet a long festering problem head on. On July 14 the CIX Board voted 4-1 with Sprint in the minority that the CIX as an association would take the responsibility for enforcing CIX member routing policy at the CIX Router. Sprint voted against the decision and in so doing apparently indicated that it favored continuing a policy of voluntary cooperation among CIX member networks. The only problem with this policy was that it seemed that virtually none of the national providers was willing to insist that CIX routing be given only to those service providers whom they connected and who joined the CIX. As a result network service providers received full CIX connectivity whether or not they paid the 10,000 dollar a year fee to join. Not surprisingly this angered an increasing number of service providers who did join. We report that according to the perceptions of some providers the pre July 14 policy seems to have been helping Sprint and ANS and hurting PSI and UUNET. We look at the NAPs as alternate interconnect points that are in the hands of two RBOCs, one IXC and one CAP. The NAP environment is open to any kind of bilateral agreements including settlements that people connecting there wish to make. It is hard to imagine installing the equipment necessary to use for settlements without also, at some future point, using it for measured usage pricing. The phone companies are all interested mightily in measured usage pricing for internet services. We know people at Sprint, MCI and Ameritech who are expressing strong interest in measured usage charging. Ameritech pricing for the Chicago NAP looks so low that some suspect it is designed to make the NAP a more interesting place to interconnect than the CIX. The CIX on the other hand with is prohibition against settlements and injunction that all members shall carry all other member's traffic is antithetical to the telco NAP point of view. The CIX in general is regarded as the internet's best hope against th e imposition of measured usage charging -- something that chapter five 'Financial Issues' of the new National Academy of Sciences study Realizing the Information Future says is the only reasonable way to evolve the network. There is strong reason to suspect that big telephone companies would like to see the CIX die rather than doubling in number of members as it has in the last four months. After all with mid to high priced PCs now capable of unmeasured voice conversations v ia the internet to practically any where in the world, the boundaries between the internet and some key telco lines of business are looking quite permeable. The impact of the routing filters will be impossible to judge until they go up. Non CIX members will likely find their connectivity impaired to some degree - how seriously can only be judged on a case-by-case basis. There is talk of the CIX implementing routing through an SMDS cloud between now and the end of the year, Some folk (Sprint among them) seem to think that the cloud will render the CIX filters meaningless. This would be true unless the CIX installed and controlled its own routers at the borders to the cloud. In addition to hoping for the survival of the CIX, those interested in the preservation of an internet capability for unmeasured usage pricing would do well to look at the implementation of SMDS and ATM as transport technologies that lend themselves to th e imposition of measured usage in the way that a private T-1 or T-3 leased line does not. We should take care to see that the telcos do not discontinue the sale of T-1 and T-3 clear channel leased lines at reasonable prices. For if the telcos bought out BBN, UUNET, and PSI and then began a rapid implementation of measured usage charging, the rest of us could set up Internet Two in about two weeks -- if --it were still possible to buy access to unmeasured, clear channel T-1 an d T-3 leased lines. The CIX has just over 90 days to emerge from adolescence into adulthood by successfully establishing the worth of its services in the way in which its routing policies are implemented. We fervently hope it succeeds. However when it takes an expansivist p oint of view that anyone passing IP packets in any form including non profits and universities are networks and must join at the full cost of $10,000 this year, it flies in the face of what many on the net who are its friends seem willing to accept.


We conclude the discussion of the activities of the ISF. We summarize the other networks - Sovam, Glasnet, Sprint Networks, and Radio MSU. Finally we include a detailed conclusion in which we analyze the considerable power in the hands of the Internatio nal Science Foundation as it impacts both science and telecommunications policy in Russia. Coming in September - a two installment section on network users.


We publish the transcript of a nearly two hour long taped recorded interview with the ISF's most important policy director in Russia. (Drs Goldfarb and Valeri Soifer were the two emigre Russian scientists who convinced George Soros to create the Internati onal Science Foundation.) Goldfarb answers questions we raised in the first half of our Russian report. He also presents his views on the Moscow backbone controversy and on ISF relations with Relcom. Finally he talks in detail about the ISF's just annou nced plans for CIVIC networks in Iaroslavl and possibly at some future point, the 30 largest cities in Russia. To our knowledge this is the first time that Goldfarb has been interviewed at such length.