Although I support both Peter Daws and Peter Kirsteins concerns about European connectivity and the possible need for European traffic to fly via the US, I like to point out that it is not the INIX, but the providers who are responsible for payload traffic exchange. As indicated in the statement, a possible and likely future requirement for an INIX, but than it will probably change name, will be to be part of a distributed GIX. The associated requirements for that are currently under investigation in the D-GIX developments activities. Kees Neggers ==> From: P.Kirstein@cs.ucl.ac.uk
While I welcome this paper I think that it is ill judged that a INIX does not also require connection to at least one European D-GIX. As a result an INIX could send all their european traffic via the US.
Peter Dawe
PIPEX
I agree with Peter, it would be most unfortunate if Europe-Europe traffic was sent via a GIX in the US, and traversed the Atlantic via any research link. I note that there are recommendations that this not happen, but I am not sure that this is strong enough. For example, because of reciprocity, the UK-US network provision may agree to accept traffic from the US GIX; it may not plan to accept traffic from a European operator who has got to that GIX via another Europe-US route.
Peter Kirstein
Peter Dawe, Managing Director PIPEX Ltd.Email: peter@unipalm.co.uk Director Unipalm Limited, Voice: +44(0)223 250100 216 The Science Park, Milton Road Fax: +44(0)223 250101 Cambridge, England, CB4 4WA