In your letter dated Wed, 14 Jul 2010 20:26:47 +0930 you wrote: >I'm a bit confused by that. My understanding of NUD was that it's main >function is to ensure that existing entries in the neighbor cache are >valid. If NUD fails, then the entry is removed from the neighbor cache >so that next time that destination IPv6 address is sent to, the absence >of a neighbor cache entry would trigger an NS/NA transaction. > >So if ND NS/NA transactions aren't performed on these point-to-point >links, wouldn't that mean there is nothing for NUD to validate? What >would be the target address of the NUD probes if they can take place >without ND NS/NA transactions priming the neighbor cache with entries?
I agree with that. I tried for a while to skip the ND step and only do NUD, but it doesn't fit within the framework for RFC-4861. The separation could be useful for 6to4. For point-to-point links, which can do multicast trivially, there is no excuse for not doing full ND (unless it is a link between two routers that actually use hellos in the routing protocol to determine whether the other side is alive). -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------