[11527] in Commercialization & Privatization of the Internet
Questions about ANS's newest T-3 customers
daemon@ATHENA.MIT.EDU (Gordon Cook)
Tue Apr 5 22:16:22 1994
From: cook@path.net (Gordon Cook)
Date: Tue, 5 Apr 1994 19:50:16 GMT
To: com-priv@psi.com
Received nearly four pounds of documents from NSF today. From what I can
assimilate so far I have a better idea of why the cost of the NSFnet "service"
continues to grow so substantially.
In Nov 1991 two T-3 connections were sole sourced to ANS for $275,000 EACH per
year to connect NASA Aimes and NASA Goddard to NSFnet ANSnet at t-3. (Add in
the local loop cost and you get roughly $770,000 as a recuring annual charge for
these connections). DARPA (ARPA now) was added at $250,000 per year in
Washington in late 92 or early 93. (DARPA local loop add on the the 250 grand
was cheap....only about $95,000 a year) MERIT's (read ANS's) price for
connecting Sandia and Phillips was a pending $586,000 a year in September 1993.
Since then I have seen an ANS backbone map saying that an Albequerque (Sp?) New
Mexico node is open. So it would appear that 4 new backbone nodes (with 5
customers) have been purchased from public monies at an average cost of more
than $300,000 per year (since Nov 1991). This when, as far as I know, ANS has
been unable to sell a commercial T-3 connection to a corporation that, unlike
the Feds, has a bottom line to worry about.
What I'd like to know is do each of these agencies get interregional
connectivity money to connect to the NAPs when and *if* the NSF contribution to
ANS for backbone services is ever halted?? (The protests underway should get
Merit a THIRD 12 month extension.) And what will the interconnect policies at
the NAPs be with MCI's vBNS if *THAT* travesty is allowed to stand? IE will a
commercial customer get connectivity to these NASA and DOE and DARPA sites as
well? Given the continued protests of ESnet, maybe the only thing DOE can do
now is hang its labs of the NSF service? Perhaps it will be the same for NASA
and DARPA too? Outside of the view of all but the most seasoned network
techies, the NSF network empire continues to grow. What happened to the ANS
press releases annnouncing these well-heeled gov't customers?
PS. I'll be off line til LATE tomorrow night.
PPS. Can any reasonable estimate be made of what PSI, Sprint or UUNET would
charge for this type of T-3 connection?? Or is such a question hopeless since the connect is to a unique gov't paid for infrastructure that according
to the fiction found in Linda Sundro's report "anyone" can connect
to on the same grounds granted ANS?