The necessity of 4-over-6 stateless address sharing mech.pptVIP

The necessity of 4-over-6 stateless address sharing mechanism Satoru Matsushima Jie Jiao Chunfa Sun * Problem space matrix based on the guideline* Crossing IPv4 Island IPv6-Only Core Network Stateful RFC5571(L2TP) DS-Lite(draft) Stateless RFC5969(6rd) 4rd (to be STD) (draft-despres-intarea-4rd) ? ? ? *: RFC6180: Guidelines for Using IPv6 Transition Mechanisms  during IPv6 Deployment Our target * Trade-off between stateful and stateless in IPv6-only core network scenario Stateful Stateless Address sharing efficiency ? Port resource flexibility ? NAT Logging ? Routing optimization ? GW redundancy ? GW load-sharing ? * * NAT Logging Considerations Using fixed NAT rules and IPv4 users can be directly identified by means of their IPv6 address. Stateful Stateless Users identified by a dynamic address and port “NAT Log” Users identified by pre-assigned static address and port-range Logging for every dynamic NAT mapping is needed No need of NAT logging Solutions Characteristics Operation * Routing Optimization: Stateful IPv4 Internet Gateway (In Prague) IPv6-only Network CE2 (in Tokyo) CE1 (in Tokyo) The packets from CE1 in Tokyo have to go through the gateway even in Prague to reach CE2 in Tokyo. Hub Spoke topology is the only choice: ISPs can use stateful solution when it’s not necessary to care about the network latency. * Routing Optimization : Stateless IPv4 Internet 4rd BR (In Prague) IPv6-only Network 4rd CE2 (in Tokyo) 4rd CE1 (in Tokyo) The packets from CE1 can reach CE2 directly. ? Achieving a optimized path. Mesh connectivity solution can be achieved on the optimized path when the communication occurs between CE1 and CE2. Our strategy of v4 to v6 transition Broadband Existing (IPv4-Only) New (IPv6-Only) 6-over-4 (2010-) 4-over-6 (2011-) * Consideration experiences: Business planning point of view Q: Total CAPEX and OPEX of a transition system T: Total number of serving customers in a transition system Comparing S value between all of solutio