Re: [DNA] Question to DT about TSLLAO and Negotiation-freeDeterministic Fast RA

[Erik Nordmark <erik.nordmark@sun.com>]

Greg Daley wrote:
> Dear DT,

Did you intend to send this on the DT list or the WG list?

> Was there a reason that Tentative Source Link-Layer Address
> options were recommended for acting as the sequence
> ordering mechanism in unnegotiated detfastra?
> 
> I would have thought that the inbound Router Solicitation's
> source address would work just as well to xor against the hash
> (probably in reverse or reverse byte order, to be less
> predictable) to get the ordering.

Since you can't assume that your link-local address is unique should you 
have moved, you need to run the optimistic DAD procedure.
While doing so, the RS must have the unspecified source address.
Hence all RSes from different hosts will all look identical, except for 
a TSLLAO.

    Erik