I have just published part one of a two part issue on the economics of peering and transit and IX's. Included is a two part interview with Bill Woodcock who elaborates in great detail on the peering and transit methodology that he has been presenting at the last couple of Nanog meetings. Also included are four tables with January 31 2002 Netflow data from Zocalo. The tables are generated with software done for Zocalo at Agilent by Alex Tudor. Bill goes through each table and explains the meaning of the data. This includes an explanation of Synthetic Path Analysis in tables 3 and 4 which as I understand it is really the first published explanation of this approach. The interviews with Bill explain over all how ISPs can begun to mold tools into a bandwidth cost management system.
My November issue also includes a long and detailed over view of the nasty consequences of the current Tier One oligopoly. The overview is an essay by Farooq Hussain in which he identifies the Internet Core backbones (see December 5, 2001 announcement of Equinix's Internet Core Service) as UUNET, Level 3, Qwest, ATT, Sprint, Cable & Wireless and Genuity. Andrew Odlyzko adds comments.
Thanks to Zocalo I am running a private mail list where discussion of the Woodcock and related methodology, exchanges, peering, bandwidth cost and the like is on going. In addition to Bill Woodcock active participants include Avi Freedman, Phil Weller (Fast net CTO), Mike Hughes, Alex Tudor, Stephen Stuart, Phillip Smith, Farooq Hussain, Andrew Odlyzko and Keith Mitchell. If there are other Peering Coordinators here (or folk playing that role) who would like to contribute to the discussion (warning ....it is for publication) please email woody@pchnet and email@example.com stating the nature of you interest. Results of these discussions and other data will come out round about October 1.
For the table of contents, complete introduction and Andrew Odlyzko's comments please visit