On 8/16/10 4:36 AM, Ole Troan wrote:
> Hemant,
> 
>>> I don't understand either. Why is it an issue for a sender node
>>> to
>> transmit a packet on the link-layer as a unicast message, if its
>> known there is only one receiver. I've not seen a single valid
>> >argument and so its fine, one is entitled for their opinions.
>> 
>> This is the email I sent out to Fred on 8/3/2010 in regards to
>> your draft - in double quotes below.  See my one reason for why
>> your document's rule will break MLDv2 protocol.  Ole agrees with me
>> that yes, MLDv2 sniffing will not work with your document's rule.
>> Further since RFC 4862 uses MLDv2 for ND control messages such as
>> an NS(DAD).  So now the router node also fails to get an NS(DAD).
>> My summary is that IPv6 address auto configuration in RFC 4862 and
>> RFC 4861 ND control is so tied to MLDv2 and RFC 3810 that if MLDv2
>> breaks for L2 sniffing as mentioned below, we just broke other IPv6
>> control.
> 
> with regards to "Ole agrees...". sure, but you didn't paste the rest
> of my argument. which is that you don't need MLD snooping for what is
> a unicast message on L2. after all this is only going to be sent to a
> single receiver.

Correct.  The MLD snooping functionality only looks at L3 information if
the L2 destination address is a multicast address.  In this case, L2 has
a unicast address and the MLD snooping function will never see the
packet (it will be forwarded using standard L2 logic).

> 
> how do you think MLD applies in this case?

I agree.  It shouldn't.

> 
> (apologies to Suresh for hijacking this thread for a L2 unicast draft
> issue).

+1.

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

Reply via email to