Jeff, On Mar 4, 2013, at 10:13 AM, Jeff Wheeler <[email protected]> wrote: > Dino's assertion that EID block registration must be performed at no > more than the cost of performing the service is myopic.
Just for clarification: I believe it was I that mentioned this in my strawman requirements for EID block management, not Dino. I understand what you are saying but believe you're trying to put the cart before the horse. As things stand now, their is little knowledge of LISP so you get into the same chicken-or-egg problem we've had with IPv6. For an experiment, I believe you need to make resources as easy to get as possible in order to minimize barriers to entry. Every barrier, be it price or legalese or justification hoops (or even figuring out where to go to get EIDs), reduces the likelihood people will be interested in playing. Hence my suggestion the allocation/registration service be no more than cost. Remember that even IPv4 addresses were given away by the millions for free back in the day. It took over a decade before user fees were required. Hopefully LISP will grow more quickly than the Internet did. Perhaps a more interesting question is what services will the EID allocator/registry need to provide (in addition to allocating EIDs and keeping registration data). Two services mentioned so far are: 1) whois, to lookup the registration data 2) reverse dns, so people can map EIDs into domain names Providing these services has a cost that correlates with a committed service level (e.g., 5 9s service costs much more than 1 9s service). What level of service is acceptable for the LISP experiment? Regards, -drc _______________________________________________ lisp mailing list [email protected] https://www.ietf.org/mailman/listinfo/lisp
