Re: [DNA] No FastRA unless Landmark? (draft-pentland-dna-protocol-00)

[Syam Madanapalli <syam@samsung.com>]

Hi Greg,


> Hi James,
>
> James Kempf wrote:
>>>The FastRA method should have been a seperate draft, unless we are going
>>
>> to
>>
>>>have a single DNA solution.
>>>
>>
>>
>> Right, so what is planned on this? Is there going to be a discussion 
>> about
>> whether more than one solution is needed? Whether people want more than 
>> one?
>> Currently, I believe we have two.
>
>
> I believed that at the last IETF the rough consensus was that
> there will eventually be one solution, which must be selected
> by the WG (with expertise from the DT to assist).

This is the best way to goforward and avoid to have multiple solutions
at the same time, which causes interoperability problems and burden on
implementers.

>
> There's only one solutions document on the charter of the WG.
>
> So I guess I don't see the need for a separate 'Fast RA'
> document at the moment.

This is true, as long as we have a single solution. I hope
we will not rework in the future to come up with a new
solution for DNA where we would also require the FastRA.

>
> That's not to say that draft-pentland-dna-protocol is the
> complete solution just because Hash Based FastRA is defined there
> (for example).

I agree with you and WG will decide. But I see two proposals
1. Landmark based and 2. Linkid based, both are from Design
team. I think somebody from the design team can put the pros
and cons of each method on WG mailing list, so that people
will have some idea before the WG meeting in Paris

>
> The Fast RA scheme is sound regardless of whether landmarks or
> completeRA are used.

Indeed.

And the idea of limiting the no. of FASTRAs to FASTRA_THRESHOLD (3)
is a good addition, hoping that no other adverse implications

>
>
> Greg
>
>