Hi Pascal,
On Oct 13, 2009, at 4:55 AM, Pascal Thubert (pthubert) wrote:
In route over, the only address(es) that the node uses at the moment
is
that of its router(s). The address is found in the RA source MAC and
the
neighbor relationship is maintain using NR/NC messages. The draft does
not attempt to shortcut for P2P - then again the model is very similar
to 802.11 infra mode. There's discussion at RPL about line of sight
P2P,
though, and it appears that we could add something like multicast
NA(override) when the address is obtained (and renewed).
I'm mostly interested in route-over, so I'll limit my discussion to
that. 1-hop P2P is something we should support - there are scenarios
where two nearby nodes may want to communicate without any additional
infrastructure.
I think a bit of my concern is the lack of clarity in the role that
6lowpan-nd plays in the router-router interface as well. For example,
what is the address resolution mechanism when communicating with next-
hop nodes - is it NR/NC? NS/NA? Something specific to the routing
protocol?
Address resolution is not used/needed/required when there is a router
between a node and its target. ICMP unreachable is your friend there.
Since the draft always goes through the router, only the router needs
actual (un)reachability testing.
The draft uses NR/NC messages to cover the node to router hop and ICMP
unreachable for all other destinations.
NR/NC only cover link-local addresses. As defined, the router can
only verify reachability to a host using its link-local address. How
does it test global addresses? Does it just trust that the host has
assigned the address to its interface when sending NR messages to
renew the binding?
Also, same thing here on whether or not 6lowpan-nd plays any role on
the router-router interface. As defined today, clearly it plays some
role because routers need to perform DAD. But what about NUD?
Address resolution? We discussed how NR/NC messages may be helpful
for certain routing protocol operations, but do they really replace
the need for NS/NA?
--
Jonathan Hui
_______________________________________________
6lowpan mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lowpan
