Re: [DNA] Comments on DNAv6 (draft-dnadt-dna-protocol-00.txt)

[Syam Madanapalli <syam@samsung.com>]

Hi Greg,


> Hi Subba,
> 
> Subba Reddy wrote:
> [cut]
>> 
>> 
>> Yes. 6+ ranked routers need to schedule the multicast even though token
>> is available.
> 
> I guess it would be a host configurable threshold.

You mean Router configurable threshold?

> 
> In general, I can't see a use for more than three
> Fast RAs (unless there are problems with the timing
> of the first three RAs - for example due to SEND
> signature computation).
> 
> How about we have fastra_threshold defaulting to
> FASTRA_THRESHOLD (3).

I think 3 is good enough. And anyway we can make it configurable.

> 
> if fastra_threshold < Rank (given that Rank starts
> at #0), then a Fast RA is scheduled at Rank * RASeparation.

Small correction,

 if  Rank < fastra_threshold  (given that Rank starts
 at #0), then a Fast RA is scheduled at Rank * RASeparation.

> 
> Otherwise, the token is put back in the bucket and
> a multicast RA is scheduled?
> 
> Greg
> 
>