Hello Remi,
On 07/02/2012 11:13, "Rémi Després" <[email protected]> wrote: > Hello Ole, Tetsuya-san, Wojciech, > > In a use case described in the 4rd-U draft (sec 5.3), an ISP replaces its > dual-stack routing by IPv6-only routing. > For this, independently from the number of IPv4 prefixes it has to support, it > uses only one mapping rule. > (By replacing each IPv4 route by an equivalent IPv6 route, it ensures that all > customers keep their IPv4 addresses.) > > For this to work, the 4rd-U draft has a bit that, in the hub&spoke case, > differs between CE-to-BR and BR-to-CE directions. Thus, packets sent to a CE > take different routes depending on whether sent by a CE or a BR. Would this be the "bit set in the V byte"? With MAP-T, the BR prefix(es) and the CE prefix(es) can be different. Matching on the source and destination prefix combination allows CE-CE traffic in a MAP domain, if one cares to detect such traffic. As you may recall classification of "MAP" and "non-MAP" traffic, is a topic we discussed in Beijing and in the design team, and there appeared to be little practical need for it from operators. Nothing in MAP precludes from adding that, if the need would actually be more substantial. Regards, Woj > > I don't see how the equivalent could work with the MAP documents you edited. > Is it that such a use case is out of scope for MAP? > Or did I miss something? > > Cheers, > RD > _______________________________________________ Softwires mailing list [email protected] https://www.ietf.org/mailman/listinfo/softwires
