Re: [lmap] Mirja Kühlewind's Discuss on draft-ietf-lmap-yang-11: (with DISCUSS and COMMENT)

[Alissa Cooper <[e-mail]>]

> On Mar 21, 2017, at 7:55 AM, Juergen Schoenwaelder <[e-mail]> wrote:
> 
> On Tue, Mar 21, 2017 at 09:11:09AM +0100, Mirja Kuehlewind (IETF) wrote:
>> Hi Jürgen,
>> 
>> thanks for you replies. I went back an had another look at the framework document which answered some of my questions. I guess the minimum you can do is to make the reference to RFC7594 normative (as well as the reference to I-D.ietf-lmap-information-model as Benoit mentioned in this comment). I guess without trying to implement it, I will not be able to figure out if there is anything missing or if I’m missing something and as such I will clear my discuss now.
> 
> To me, it feels a bit strange to make RFC 7594 normative

I agree, there is nothing specified in RFC 7594 that needs to be implemented in order for the YANG model to be implementable.

Alissa

> but if there
> is IESG consensus that RFC 7594 should be normative I will implement
> the change and hold my breath.
> 
>> I still think that it could be good to give some further guidance on how connectivity is established. Something like, in most cases the controller will connect the MA and the controller should make sure that it reconnects frequently based on the timeout configuration of the MA. If the MA e.g. is behind a NAT, the MA must establish the initial connection and try to reconnect when the timeout expires. Btw. is it enough to open a transport connection or do you mean by checking connectivity that there also should be some data transmitted to ensure that the controller is no only reachable but also active?
> 
> Depends to some extend on the protocol used. I think it is more than a
> TCP connection, you also need to successfully authenticate (I
> think). And then some protocols immediately become chatty - NETCONF
> has a <hello> exchange, with RESTCONF things could be silent but I
> would assume a controller will likely fetch some status information.
> (When you kid returns home, you likely ask how it is doing. ;-)
> 
> /js
> 
> -- 
> Juergen Schoenwaelder           Jacobs University Bremen gGmbH
> Phone: [phone]         Campus Ring 1 | 28759 Bremen | Germany
> Fax:   [phone]         <http://www.jacobs-university.de/>
>