On 2007-08-17T06:59:56-0700, Templin, Fred L wrote: >> How can that happen with a DHCPv6 host? RA will always precede DHCPv6 >> transactions because unless the host sees an RA with M bit set the >> host will not initiate DHCPv6. > > That doesn't make much sense; if a node doesn't hear > RAs, why wouldn't it try DHCPv6 before giving up? > > Fred > [EMAIL PROTECTED]
I agree. If you are sending RAs from your router and simultaneously need to have a node delegated a different length prefix than what is provided by the RA, then DHCPv6 is the place to do it. If there is no prefix length specified in the DHCPv6 offer then the prefix length from the RA wins, else DHCPv6 wins. I can't think of any reason off of the top of my head to do so, but why forbid the option? If you aren't sending RAs from your router (the reasons for not sending RAs notwithstanding), as Fred eludes, then the node needs to learn the prefix length along with the address via DHCPv6. If DHCPv6d authors want to make it hard to turn this knob so that admins don't inadvertently advertise inconsistent prefix lengths, that's their choice, but to forbid the option in the specification seems wrong. Regards, Mike -- Mike Oliver, KI4OFU [see complete headers for contact information] +----------------------------------------------------------------------+ | KI4OFU When all else fails...AMATEUR RADIO THRIVES! KI4OFU | +----------------------------------------------------------------------+
pgpcu5wNNItbj.pgp
Description: PGP signature
-------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
