Jinmei, > That is, the valid and preferred lifetimes of an address configured by > DHCPv6 (the "stateful" protocol) can be updated by succeeding RAs. > Does this really make sense? For example, consider the following > scenario: I personally think that it makes sense to keep this specification. The scenario you describe (address assigned by DHCPv6 gets invalidated by RA) can be relevant in cases where a DHCPv6 server infrastructure breaks down for some reason and where you quickly want to react by sending certain RAs so that you can "switch" a subnet so SLAAC. Sending RAs might then be the only way to control addresses until a new DHCPv6 server is up. This would also help in cases where rogue DHCPv6 servers assigned addresses you don't want. You can invalidate them by sending RAs in this case.
> Besides, this part of the specification seems a bit too specific about > stateful autoconfiguration, considering we are now going to separate > particular behavior on the stateful configuration part from > rfc2462bis. I understand what you mean, but we should make clear that all forms of autoconfigured addresses can be influenced by RAs. Maybe it would make sense to reword the section from ... of an autoconfigured address (i.e., one obtained via stateless or stateful address autoconfiguration) in the list of addresses... to ... of all autoconfigured addresses in the list of addresses... Do you think this wording would work while not being too specific? Christian -------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------