Re: DoCoMo Implementation Issues with CTP

Hi Hesham,

Soliman Hesham wrote:

> => Just out of curiosity, why do you expect the protocol
> to function in this scenario? If I understand you correctly
> the MN is moving from a v6 AR to a v4 only one. Even if
> CT worked in this scenario, many other things will break.
> Is there a requirement for CT to work here ?
>

The issue is separation of MN's contexts (belonging to
both IPv4 and IPv6 addresses) and the IP addresses used
in the Src, Dst fields in CTP messages. The assumption is
that routers know what IP version to use for transporting
the CTP messages. But that version (i.e., v4) does not necessarily
indicate whether the AR is dual-stacked or not to support
the contexts themselves. If an AR receives contexts it does
not support for whatever reason, it discards them.

Hope this clarifies.

-Rajeev


>
> Hesham
>
>  >
>  > > >
>  > > > > Is the confusion in whether `V' bits also refer to
>  > > > > PAR and NAR addresses ? If so, we should clarify that they
>  > > > > refer to MN's IP addresses only, and that the router address
>  > > > > must be either IPv4 or IPv6 but not both.
>  > > > >
>  > > > This description in the draft may clarify things, but how to
>  > > > identify the IP version AR in CT messages.
>  > > > I guess another V flag for AR might simply things.
>  > > >
>  > >
>  > > Perhaps this is not necessary. (See above)
>  > >
>  > I mean,
>  > In CTAR Message there is pAR IP,
>  > how can nAR know that there is IPv4 or IPv6 address of pAR,
>  > if the V flag is only for MN IP version.
>  >
>  >
>  > regards,
>  > Raghu
>  >
>  > _______________________________________________
>  > Seamoby mailing list
>  > [email protected]
>  > https://www1.ietf.org/mailman/listinfo/seamoby
>  >
>
> _______________________________________________
> Seamoby mailing list
> [email protected]
> https://www1.ietf.org/mailman/listinfo/seamoby