Excellent points Thomas. > > 5. what if the M flag is set but the host does not get any DHCPv6 > > Advertise in the initial exchange? Is it okay to fall > back to the > > RFC3736 subset? Or is it even okay to run both full RFC3315 and > > the RFC3736 subset concurrently from the beginning? > > Should be legal, but this is an unfortunate situation, since we're not > trying getting into how to deal with misconfigurations... >
Probably the correct behavior for a client is to fallback to 3736 but periodically try 3315. This is similar to what Microsoft clients have been doing for a long time ... If configured for DHCPv4, they'll attempt it for a while and if it fails, they autoconfigure an address. But, that doesn't mean they stop doing DHCPv4. In fact, the host continues to do so (albeit at a less aggressive rate). So, for a *full* featured DHCPv6 client, if it is told to run (either because of local policy or because M bit set), if SHOULD attempt stateful address configuration. After some number of attempts (perhaps only 1), it may attempt stateless (3736). If it gets stuff, great. However, in any case it should continue to send Solicit's periodically. Isn't it the case that for Stateless Autoconfiguration a host that never receives an RA for its RS will periodically send RSs? Though perhaps that's not correct because a router will periodically send RAs. Regardless of which way it works, the point is still the same ... There are periodic attempts to communicate and that should be no different for DHCPv6. - Bernie -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------