Re: [Roll] Way forward for draft-clausen-lln-rpl-experiences

[Martin Heusse <[e-mail]>]

Jean-Philippe,

you wrote:

> How often do you think that ISIS LSA should be refreshed ?
> Or OSPF opaque LSA for TE extensions be flooded ?
--> "It depends"?
I am not saying that we would need to mention in the (general) RFC when, or at which rate,
- global repairs should take place or;
- DAO should be sent out on a regular basis…
(Although this could perfectly appear in applicability, isn't it?)

My point was that the DAO mechanisms are not so clearly defined... Which may be fine: then Mukul is right and if you really need to reach everyone, then P2P RPL is there...
Right now, there can be a problem with DAOs
- if they are lost along the way (not unlikely if not using DAO-ACKs **even with L2 ACKs**, still possible otherwise) 
- if a sub-dodag moves... all routes are broken, and we don't really know how to re-establish them (although DTSN could help here, although it's a little vague)
- if an inconsistency is detected on a downward route, you'd have to ask everyone to re-send DAOs (Pascal's DAO requests would have helped of course -- this shows up on the list every now and then).

--> for opaque LSA, the information in the packet does not matter, but the flooding mechanism is robust and well defined. For DAOs, we know how the packet must look like...

> Or BFD packets ? 
> Or PCEP Keepalive when turned on ?
> 
> Very seriously … these questions apply to all protocols … and you cannot expect the specifications to give you
> hard coded values (but default whenever possible); these are topology and application specific, don't you think ?
Yes, I agree! 

Martin