>> in MAP aggregated mode you have: >> 1.1.1.0/24 -> 2001:db8:XXXX::/48 > > Actually, the so called 'aggregated mode' is original MAP-E, with IPv4 > address and PSID embedded in the EA-bits field in End user IPv6 prefix. Let > me try to explain what you have 'aggregated'. > > 1.1.1.1:01 -> 2001:db8:101::/48 (EA-bits: 0x101) > 1.1.1.1:02 -> 2001:db8:102::/48 (EA-bits: 0x102) > 1.1.1.1:03 -> 2001:db8:103::/48 (EA-bits: 0x103) > . > . > . > > The IPv4 address and port set are the same in what you called 'MAP 1:1'. > However, the corresponding IPv6 addresses are totally different. Because one > uses MAP algorithm, the other doesn't. > > > The original MAP-E achieves address sharing by embedding IPv4 suffix and PSID > into End User IPv6 prefix. But if the EA-bits is ZERO, PSID is also ZERO. > Nothing embedded in the IPv6 prefix. In this case, the so called 'MAP 1:1' > abandons the MAP algorithm and uses a per-subscriber stateful kind method to > achieve address sharing. That is NOT MAP-E. > > So I question why 'MAP-E 1:1' bears the MAP-E name.
I really don't care what we call it. what we are talking about is simply RFC2473 tunneling with extensions from RFC6346. if you want to, please suggest whatever name you like. what I'm arguing against is that we need a separate RFC for the mechanism if it uses a /48 versus if it uses a /40. cheers, Ole _______________________________________________ Softwires mailing list [email protected] https://www.ietf.org/mailman/listinfo/softwires
