On 18/10/2016 11:20, Michael Richardson wrote:
> I read draft-ietf-anima-prefix-management-01 today looking for a model for
> how to define an ASA. I find section 5, and I it is reasonable, but I feel
> that there are perhaps some critial details missing.
> I think that it's pretty important to be more verbose than you might think
> in the early ASA documents, because it will be copy and pasted.

I'm sure that's correct. I think the best way forward is to code and debug
a toy version of this ASA, which any competent Python programmer could do...
not sure it needs to be me every time. Then it would be fairly straightforward
to extract the missing details from the code and turn them into English.

(This is also the motivation for draft-carpenter-anima-asa-guidelines, so
comments on that are welcome too.)

> For instance, I think that there might be a negotiation phase after the
> discovery... that might involve a number of round trips.  I also think that
> there might be multiple negotiations, and M_WAIT, M_SYNCH or M_NEG_SYN might
> be involved, such that an aggregate prefix will be negotiated for subsequence
> division...

Yes, there would certainly be a negotiation. This is the classical GRASP
case, actually: negotiating a scarce resource. That's why I chose the
money negotiation example way back in early GRASP presentations. Slide 14
from IETF92 attached. The only difference would be that instead of asking
for $100, the initiator would ask for a /48 (say) and settle for a /52.

> ====
> I then went on the section 6.x which details with quite a lot of ideas about
> Intent.  They are interesting ideas, but I think that they do not belong
> there.

Well, Intent is part of an ASA's dependencies. Where else would we document

> I found it interesting (a self-comment on my clue) that I have no idea what
> eany of: RNC Site Gateway (RSG), Aggregation Site Gateway (ASG), Cell Site 
> Gateway
> (CSG). I'm getting it has to do with 3G/mobile IP systems.

And it shows another thing: ASAs will be defined by subject-matter experts,
not by autonomics experts.


Attachment: p14.pdf
Description: Adobe PDF document

Anima mailing list

Reply via email to