Re: [MEXT] Home Agent Switch Message (RFC 5142) and DSMIP

[Julien Laganier <julien.laganier.ietf at googlemail.com>]

On Monday 03 November 2008, Vijay Devarapalli wrote:
> Julien Laganier wrote:
> > Vijay,
> >
> >>>>> Wouldn't a new DSMIP companion document just define a new
> >>>>> mobility option to use with HA Switch and how it should be
> >>>>> used? I wouldn't think we'd revise 5142 to add DSMIP-specific
> >>>>> content to it.
> >>>>
> >>>> That will result in two documents for the same purpose. We
> >>>> should avoid that. I don't see any issue with 5142bis talking
> >>>> about the IPv4 home agent address.
> >>>
> >>> I see a possible issue with having both plain-MIPv6 and DSMIPv6
> >>> support in the same RFC, e.g. 5142bis.
> >>>
> >>> There will be MIPv6 only systems with HA switch functionality but
> >>> without DSMIP capability. Can such systems claim conformance to
> >>> 5142bis?
> >>
> >> We can add a mobility option for the carrying the HA IPv4 address,
> >> not modify the message itself. Existing implementations would
> >> become conformant to 5142bis when they get updated to support the
> >> new mobility option. This is normal, though.
> >
> > But I don't see why a MIPv6 only systems would need to support the
> > new (IPv4) option when it doesn't support DSMIPv6...
>
> 5142bis would obviously say that this new mobility option, that
> carries the IPv4 address of the home agent, MUST be implemented if
> the home agent supports DS-MIPv6. So if you are planning on a MIPv6
> HA implementation without DS-MIPv6 extensions, you can skip
> implementing the new mobility option in 5142bis.

That wasn't obvious to me. The approach is fine w/ me.

--julien
_______________________________________________
MEXT mailing list
MEXT at ietf.org
https://www.ietf.org/mailman/listinfo/mext