Brian, > > Ran and I drafted this in reaction to the earlier discussion here > about the impracticality of renumbering. We'd like comments and, above > all, contributions. For now, this list is suggested for any discussion, > but I will also mention the draft on the intarea list.
I completely agree that renumbering needs work, but I don't think that we can approach the problem by looking at all the existing devices and applications in a network and see how they can survive a renumbering event and list all the practical problems that need to be solved. I think that we should take a completely different approach. I believe that the networks, small, medium or large, that are able to easily renumber their addresses will have a competitive advantage in the long term compared to the networks that cannot renumber because they will be more agile and will adapt more easily to changes in providers and so on. The document that we need in the long term is not a list of all the problems that we have today. It is a list of guidelines on how to build an easily renumberable network. We should consider different types of networks : home network, SME network, campus network, enterprise network, small ISP network, ... and for each network define a set of services that are used inside the network such as firewalls, DNS, routers, applications, ... and for each service explain how it should react to a renumbering event. I agree with you that in most cases, this implies thu utilization of strong asset management to "automatically" build the configuration of the services. I don't see strong asset management as a limitation as it has other benefits in addition to renumbering. Based on such a document, developpers of the different services could improve their service to better support renumbering, e.g. : - libraries used by applications could check DNS TTL - applications could rely more on DNS than on IP addresses - routers could better support dhcp prefix delegation - firewalls could be configured based on information stored in DNS or other databases instead of lists of IP addresses - ... I'm convinced that with some small changes we could significantly improve the "renumberability" of services and applications. Olivier -- http://inl.info.ucl.ac.be , UCLouvain, Belgium _______________________________________________ rrg mailing list [email protected] https://www.irtf.org/mailman/listinfo/rrg
