Re: AD review of draft-ietf-rtgwg-yang-rip-04

[Robert Wilton <[e-mail]>]

So the conversion tool has worked OK on the RIP model as well, but I 
spotted a few areas where manual conversion is required (because the 
types/structure between config and state differ):

So along with the revision date, and a few FIX ME comments, the 
following few places also need to be manually tweaked/fixed:

rwilton@rwilton-lnx:~/ietf-models-to-combined/draft_modules$ pyang -f 
tree --ietf [e-mail] > 
[e-mail]
[e-mail]:1: warning: unexpected modulename 
"ietf-rip" in [e-mail], should be ietf-rip-nmda
[e-mail]:1: warning: unexpected latest revision 
"2017-06-05" in [e-mail], should be 2017-09-21
[e-mail]:740: error: unexpected keyword "type"
<- Means that config and state type differ.

[e-mail]:761: error: unexpected keyword "type"
<- Means that config and state type differ.

[e-mail]:818: error: there is already a child node 
to "interface" at [e-mail]:636 with the name 
"originate-default-route" defined at [e-mail]:731 
(at [e-mail]:141)
<- Trying to merge an "originate-default-route" leaf from the state tree 
with the "originate-default-route" container in the equivalent config tree.

RIB YANG model converted to NMDA structure attached.

Thanks,
Rob


On 20/09/2017 18:27, Alia Atlas wrote:
> As is customary, I have done my AD review of 
> draft-ietf-rtgwg-yang-rip-04. First, I would like to thank the 
> authors, Xufeng, Prateek, and Vikram, as well as the WG for their work 
> on this document.
>
> My one major issue is that this does not conform to the NMDA 
> guidelines - where augmenting -state models is not preferred.  It is 
> quite acceptable to have that in an appendix, if there are 
> implementations. I do see the shepherd's write-up indicates a partial 
> implementation exists.
> There is some tooling to help convert a model to conform to NMDA; I've 
> cc'd Rob Wilton, who was working on that.
>
> I also have some questions.
>
> 1) For the prefix-set-ref, I don't see any information about what the 
> string should contain.
>
> 2) For the route-policy-ref, I don't see any information about what 
> the string should contain.
>
> Nits:
> a) p.26:"choice auth-type-selection {
>                  description
>                    "Specify the authentication scheme.
>                     The use of the key-chain reference here is
>                     designed to align with other proposed protocol
>                     models.";"
>    Since the key-chain model is approved for RFC publication, the 
> description can be updated.
>
> Once the model conforms to the NMDA guidelines, I will be happy to 
> advance this draft to IETF Last Call.
>
> Thanks,
> Alia