Hello all, Please find attached submission on NSI inter-network topology model from John Vollbrecht. Guy -----Original Message----- From: John Vollbrecht [mailto:jrv@internet2.edu] Sent: 30 March 2010 17:52 To: Guy Roberts; Jerry Sobieski; Tomohiro Kudoh; Joan Antoni Garcia Espin Cc: John Vollbrecht Subject: submission for NSI arch Guy and all -- This took longer than I had hoped to finish. Please distribute it to the NSI group as a whole if you think it appropriate. I include Joan on this because I think (hope) he will be interested in the authorization/topology issues in section 2- especially since he has volunteered to do something with the concept of "owning". Attached is a submission for NSI protocol spec. I am not sure how it fits with what exists now, but I think it could be in a introduction or overview section. A couple notes - I did not use the term STP. I tried, but it didn't seem to fit, and I still have a problem with understanding its precise definition in different circumstances. This can be changed once we sort this out. I tried to separate the part that deals with multi-NSA infrastructures from the basic interface description. I think this works, but maybe not. I added some description of the interaction between NSA authorization and transport topology. I hope this is a good start and can be useful in separating discussions of topology and authorization which probably should happen in a limited way before releasing the recommendation I included some high level description of paths. I think (hope) this is good intro to some of the stuff Jerry has written. As a interesting note, in the path discussion it describes path as Path == [ingress (-egress, ingress)* - egress] It is interesting to note that the term (-egress, ingress) might be considered a Point or edge (or even a link) between networks I am most interested in reactions and comments. Thanks John