Le 1 août 2011 à 16:31, Jan Zorz @ go6.si a écrit : > On 8/1/11 4:22 PM, Rémi Després wrote: >> >> Le 1 août 2011 à 15:36, Rajiv Asati (rajiva) a écrit : >> >>> ... Interesting enough, the static port-set is one of the reasons why many >>> find 4v6 being so useful. >> >> Indeed: operation simplicity, scalability, possible direct CE-CE paths. >> Very legitimate. > > Let me repeat, what was "thrown" at us, when writing ymbk-aplusp: > > "what happens, if customer behind CPE that shares IPv4 address runs out of > ports? Does this mean that all his additional traffic from this point does > not go anywhere? To end customer, this looks like broken internet > connectivity."
Thanks for the quotation. What was "thrown" to you doesn't seem right, and needs to be challenged. Answer to the question: - The customer doesn't run out of ports because it has an exclusive private IPv4 address. - If the CPE NAT44 runs out of ports, is just does what it does today in the same situation. In addition, the probability of the CPE running out of IPv4 ports in a dual-stack-service site having 4K ports needs not be higher than in an IPv4-only site having 64K ports. > Don't get me wrong, just repeating what we needed to solve in order to move > on from that point. I think I don't get you wrong, and actually I believe your pioneer work on the subject has been quite useful. It is just that what you felt obliged to solve isn't in reality needed in a significant number of legitimate use cases. > Basically, to solve that we needed to introduce states at some point. > If we can solve this issue on stateless solution, I'm all for it. See above. Cheers, RD _______________________________________________ Softwires mailing list [email protected] https://www.ietf.org/mailman/listinfo/softwires
