Re: [Mip4] New revision for rfc3344bis; UDP inconsistency and candidate solution

[Sami Vaarala <sami.vaarala at iki.fi>]

Charles E. Perkins wrote:
> Hello Ahmad and Sami,
>
> As far as I know, the proposed wording change is
> backward compatible.
>
> Regarding port forwarding NATs, I hope there are none
> that change the destination port!  The home agent can't
> be listening to all the possible destination ports and
> treating all incoming messages to all ports as Mobile IP
> messages.
>
> I can see that a NAT box might change the source port
> number, and the wording changes I proposed would not
> affect that.

There are HAs that can listen on other port(s) beside 434.
Operation on these other ports is not directly covered by
the specification as such, but it would be preferable that
the wording "just work" in that case, too.

In particular, it would be bad if an implementor decided
to support running a HA on a non-standard port, receiving
requests on some port X, but still sending responses using
port 443 as source port.  Even though that might work in
some cases, it would likely fail with NAT devices and other
stateful routers; thus it is important for the HA to copy
ports exactly from the RRQ (just swapping them).

That being said, if we want to *preclude* using other ports
in the base specification, then this change is a no-op.

Best,

-Sami

--
Mip4 mailing list: Mip4 at ietf.org
   Web interface: https://www1.ietf.org/mailman/listinfo/mip4
    Charter page: http://www.ietf.org/html.charters/mip4-charter.html
Supplemental site: http://www.mip4.org/