Folks, I have not read this draft yet and I am also not familiar with any IKE. But two things caught my attention in Pasi's presentation today. They were: (a) He was looking for a link model for his IKEv2 for IPv6. (b) He said the VPN concentrator in his IKEv2 needed to support 40K-50K clients and it was not possible to have so many disparate network interfaces on the VPN concentration router. Having heard just these two things, I did suggest in the meeting that an aggregation router model fit his link model. As Erik has explained in his email, what I call as an aggregation router is what IETF community calls as a point-to-point link model of large number of clients attached to a single router. I met Pasi too and explained to him on paper what I called as an aggregation router and he explained to me what was IKE2. I worked out an aggregation router link model for his VPN concentrator. Then we worked out a new call signaling flow so that a single IPv6 address can be assigned to the client by the VPN concentrator. We left with the thought that no gotchas were left in the new design. He thinks an Appendix can be added to his draft to show this model we developed today. We'll wait for him to reply. Hemant
-------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------