Re: [DNA-BOF] Do we need a separate DAD Problem Statement? (was Re:About Charter, DAD, and SEND)

[Youn-Hee Han <yhhan@sait.samsung.co.kr>]

Hi Greg,

Greg wrote:
> >>I agree that it is important to be able to do DAD (optimized)
> >>before DNA has occurred. Maybe we need to say:
> >>
> >>"3.N+1  Simultaneous operation with DNA
> >>
> >>  It SHOULD be possible to use DAD optimization procedures
> >>  to ensure availability of link-local addresses before DNA
> >>  has been completed.  This would allow DNA procedures to
> >>  be undertaken without needing to wait for [RFC-2462] style
> >>  DAD completion."
> >>
> >>I think this separates the optimism (which may provide a solution)
> >>from the goal.
> > 
> > 
> > Does "DAD optimization" include the validity check of previous 
> > IP address on a new link? IMO, we had better confine "DAD 
> > optimization" only to the matter reducing the delay time.
> > 
> > We should not be confused between "the validity check of
> > previous IP address on new link" and "the reduction of 
> > delay incurred by duplicate address detection", although the
> > two schemes help send and receive packets immediately after
> > network attachment.
> 
> I think that if we're checking the validity of a _global_ address
> which has been previously used on another link (through DNA),
> then that shouldn't interfere with DAD optimization.

I completely agree.

> It may be valuable to check previous link-local address or get a new 
> valid link-local address with an Optimized DAD scheme.  I don't
> mind which.
> 
> We are talking about the validation occurring on the new link,
> performed by the host (or by the host's signalling request sent on
> the new link), though, aren't we?

What I am concerned about is that the word "optimization" includes
the validity check of previous IP address on new link.
IMO, the validity check should not be in scope of "optimization" 
works. It is just one of the matters of network attachment. 
Do you agree? If you do not agree, we may update the meaning
of "DAD optimization" in our charter.

To check IP layer connection (network attachment), a node 
should check wheter
1) it has reachable default router.
2) it has valid IP address.

I agree that the check of 2) and the (optimistic) use of 
previous IP address can be done before DNA has 
been completed.
But, I am not sure that DAD optimization (new IP address
configuration while reducing the delay) can be done before
DNA has been completed. Do you think it can?

Youn-Hee