On 2005-10-19, Christian Vogt wrote:
> 
> "The initial Neighbor Solicitation MUST be transmitted as early as
> possible after the Optimistic Address has been flagged as 'Optimistic',
> but it MUST NOT violate any delays or rate limitations set forth by
> RFC2461 or RFC2462.

Actually, I disagree.  In order to provide timely address configuration,
we MUST violate these delays.  

As I understand it, until we've performed the MLD report, we can't
reliably receive NSes ... and thus we may not be able to send our
NA(O=0) and can't respond to communications!

Also, the longer we delay the NS, the longer we have to wait to discover
a collision.  This is a big disadvantage to the arriving node, because
in the case of a collision it's not getting any traffic.  It's also
a small disadvantage to the collidee, because it may get traffic 
which was meant for the arriving ON.

I think L2 collision avoidance is best left to L2 in any case.
Elimination of these delays was proposed by some "Fast RS" draft
or another at some point, although I don't think I ever wrote up
the Soft vs. Hard Handover draft I'd been intending to ...

-----Nick

--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to