On Tue, 2005-05-24 at 10:04, Soliman, Hesham wrote:
> I think this draft is now ready for the IESG.
I have one issue with APPENDIX A. The direct mention of the on-link
assumption was removed from the appendix (from the second bullet item
1), but it is still implied by the text that remains. The text is:
1) If no Router Advertisement is received on any interfaces, a
multihomed host will have no way of knowing which interface to
send packets out on, even for on-link destinations.
One possible approach for a multihomed node would be to attempt
to perform address resolution on all interfaces, a step
involving significant complexity.
Why would a multihomed host ever attempt to perform any address
resolution on any interface unless it assumed that the destination was
on-link on one of those interfaces? It should not make this assumption
at all. While the text is accurate in stating that this involves
significant complexity, it should not suggest that doing on-link address
resolution for a node to which it has no route is a "possible
approach". If you're multihomed and you have no route to a destination,
then send back an ICMP destination unreachable. It should be as simple
as that.
If the destination is a link-local address, then that's a different
matter entirely. In that case, it's a matter of scope ambiguity. The
text may have been addressing this particular case, but it's not clear
to me.
-Seb
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
