Re: [DNA-BOF] About DAD optimization solutions space

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

Hi Pekka,


Pekka Nikander wrote:
> [If anyone feels I should limit my posting rate to DNA,
>  please state so.  OTOH, I think that the current issues
>  are converging fairly fast.]

You're being helpful, so please don't stop posting.
If we can't keep up we'll tell you.

> Greg Daley wrote:
> 
>> For the DAD optimization especially, since there are
>> existing solutions, a set of goals may be sufficient.
>>
>> I was actually thinking that we could aim at producing
>> one optimization scheme as a 'first' standard, though.
>> If we're going to choose one to standardize first though,
>> we'd need to clearly demonstrate its sufficiency with
>> regards the the goals and requirements.
> 
> 
> To me, it looks likely that we may need two different
> DAD optimization solutions: one for probabilistic, stateless
> addresses, and another for statuful, pre-reserved addresses.
> Preferably, these two should work at the same time, if needed.
> The former category has the benefit of working with today's
> routers and seamlessly with SEND CGAs, the latter has the
> benefit of supporting non-probabilistic addresses and perhaps
> working faster when both the routers and nodes support it.

I think this is a distinct possibility (two solutions), if
they have different strengths which may be exploited.

I'd really prefer to get at least one solution out quickly,
especially if we need to refine or select the stateful system
for robustness, or wide applicability.

We'll have to be careful to define what the strengths and
weaknesses of each system are if we try for two solutions.
IETF doesn't typically produce two solutions where one will do.
Well it does, but I don't think it aims to.

Greg