George Tsirtsis escribió:
On Mon, Mar 31, 2008 at 2:04 PM, marcelo bagnulo <marcelo@it.uc3m.es> wrote:George Tsirtsis escribió:On Fri, Mar 28, 2008 at 6:15 PM, marcelo bagnulo <marcelo@it.uc3m.es> wrote:> >> Hi, >> >> during the v6ops meeting, Dave (and maybe others) brought up an issue >> about the support for other protocols, like MIPv6, SCTP, DCCP and others. >> >> Curerently the draft is phrased as follows: >> >> I6: MIPv6 support >> >> The translation mechanism SHOULD not prevent MIPv6 Route Optimization >> when the CN is a v4-only node >> >> > > GT> I am confused about this one. MIP6 RO requires the CN to > participate in MIP6 signaling for the Return Routability tests. How > would an IPv4 only node participate in such an IPv6 specific > signaling?. This makes no sense to me. > Well, what people have suggested is that you can perform the RO to the NAT64 box (i.e. the NAT64 perfomrs CN operations on behalf of the v4 node)GT> I would think that in most cases the actual routing optimization benefit from such an operation will be zero. IMO we should not even bother.
not only that, but we should also make sure that the HoT/HoTI and CoT/CoTI fly through the same NATPT box (so it can intercept them) which is far from obvious, i guess, so this would work in a limited number of cases
Regards, marcelo
> I was also thinking if any other MIP related scenario is worth > considering here, e.g., MIP HA inside vs outside the NAT64. MIPv4 and > MIPv6 protocols, however, are entirely incompatible so MIPv4 to MIPv6 > translation is not a realistic option as far as I can tell. My > inclination is to just forget about Mobile IP and all its versions and > denominations for now. > > i am ok with that tooGT> OK :-)Regards, marcelo > George > >