Re: [DNA] draft-ietf-dna-frd-00.txt clarification for RA trigger

[Greg Daley <greg.daley@eng.monash.edu.au>]

Hi Kent,

Kent Leung (kleung) wrote:
>  
> Hi folks.  Just joined the alias.  Hope this has not been discussed already.
>  
> This section says that the PoA can send RS to trigger RA from the AR.  
> What's not clear is what source link-layer address is put into the 
> TSLLAO option (maybe it's implicit)?  I'm assuming that the source 
> link-layer address of the RS is the PoA, and the address in TSLLAO is 
> the host so the RA can be unicasted directly to the host.

Yes, I think that this is the intention.

> 3.1  RA Triggering
> 
>    There are two ways for such a remote Link Up event notification.  We
>    may use the MIES (Media Independent Event Service) defined in IEEE
>    802.21 [21] or RS with TSLLAO (Tentative Source Link-Layer Address
>    Option) [14].
> 
> In the 802.11 example below, no mention of RS method.  In this case, the
 > AP will send RS with source MAC address being AP's and the
 > TSLLAO containing the host.
> 

Yes, I think so.


> 5.2.1  802.11
> 
>    In 802.11 Wireless LAN technology, when a new host arrives at an
>    AP(Access Point), it should associate with the AP.  The host sends an
>    Association Request Message with its MAC address.  Then the AP sends
>    an Association Response Message to grant association.
> 
>    As soon as association is made, the AP sends a cached RA to the host
>    in an unicast 802.11 frame with the MAC address from the Association
>    Request message.  The host receives the unicast RA just after
>    association is made, which is the earliest possible time in current
>    standard.
>  
> 
> What is the status of  draft-daley-ipv6-tsllao-02.txt?
 > Is this being adopted in IPv6 or DNA WG?

It may be adopted in DNA working group.
That seemed to be the feel of the room at the session.

That said, the draft wasn't presented in a regular way
(not advertised in the agenda, and I was actually surprised
that Suresh asked me to make a presentation on-the-fly).
So we definitely need to make an on-list consensus call
as to whether the document should be adopted in DNA as a
WG document.

Greg