In einer eMail vom 11.01.2010 19:43:35 Westeuropäische Normalzeit schreibt [email protected]:
While your subject line piques my interest, there is not enough context either in the original message you are quoting or in your responding text below for me to understand what you are claiming. I would suggest that you more formally lay out your point of view. Regards, Eliot With classic DV-complying BGP you receive reachability information from some router precisely then when that router allows you to use him for transit. Otherwise he won't give you that information. Hence this sort of policy is implemented implicitly. With LISP+ALT you have to pull the needed ETR information from some ALT-entity and I can hardly imagine neither that the requested RLOC information is stored there, combined with who is and who is not allowed to get it. I was told several times that I would not understand the internet economics, because TARA would use Dijkstra to determine the best next hop (based on some combined map of several and different zooms). I could only defend TARA by saying, well then any restriction "who is allowed to use which link" needs to be communicated somehow explicitly. But obviously LISP is in no better shape. Just the opposite. I think, that LISP is even in a worse shape wrt this restriction issue. And this is not the only issue: see the horrible overlapping EID discussion - especially as there is no need at all to propagate any single prefix. Heiner
_______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
