Re: [Netconf] [yang-doctors] Yangdoctors last call review of draft-ietf-netconf-netconf-event-notifications-16

["Eric Voit (evoit)" <[e-mail]>]

Hi Reshad,

Thanks for the review.   Some thoughts in-line...

> From: Reshad Rahman, January 9, 2019 1:49 PM
> 
> Don't know why newlines aren't showing up in the email, if you want a nicer
> view (also a slight revision):
> https://datatracker.ietf.org/doc/review-ietf-netconf-netconf-event-
> notifications-16-yangdoctors-lc-rahman-2019-01-09/
> 
> On 2019-01-09, 11:34 AM, "yang-doctors on behalf of Reshad Rahman" <yang-
> [e-mail] on behalf of [e-mail]> wrote:
> 
>     Reviewer: Reshad Rahman
>     Review result: Ready with Issues
> 
>     YANG Doctor review of draft-ietf-netconf-netconf-event-notifications-16 (by
>     Reshad Rahman)
> 
>     No YANG module in this draft but there are numerous examples based on
> the YANG
>     modules in draft-ietf-netconf-subscribed-notifications and
>     draft-ietf-netconf-yang-push. Validation of the examples was done with
> yanglint
>     0.16.59.
> 
>     The major issues are related to the examples, although in 2) below  the
>     root-cause might be with the YANG module in
>     draft-ietf-netconf-subscribed-notifications  and in 3) there is an issue which
>     has already been reported on draft-ietf-netconf-yang-push: 
>     1) The  establish-subscription request in Figure 3 fails yanglint validation, the
>     "stream" element has to be moved after the "stream-xpath-filter" element,
>      that  fixed the following issue: err : Invalid order of elements
>     "stream-xpath-filter" and "stream".
>     (/ietf-subscribed-notifications:establish-subscription/stream-xpath-filter) 
Done.

>    2) The establish-subscription error response in Figure 5 fails yanglint
>     validation, the "id" element is missing. This is because
>     draft-ietf-netconf-subscribed-notifications has leaf id as mandatory in
>     establish-subscription output, either we remove the mandatory or we
>     reserve a
>     value as invalid id or we add err : Missing required element "id" in
>     "establish-subscription".
>     (/ietf-subscribed-notifications:establish-subscription)
The id is mandatory for the <establish-subscription> RPC response of that RPC.  It is not needed as part of the general <rpc-error> response.   So I believe we are good here.  (Note: Yanglint would need to be upgraded to do fairly complex RPC correlation to be able to automatically cover such issue types.)

>     3) The  modify-subscription request in Figure 7 fails yanglint validation 
>     because of missing "datastore" element. err : When condition
>     "../ietf-subscribed-notifications:target/ietf-yang-push:datastore" not
>     satisfied.
>     (/ietf-subscribed-notifications:modify-subscription/ietf-yang-push:periodic)
>     There are 2 causes here: missing "datastore" element and incorrect when
>     statements in draft-ietf-netconf-yang-push (as reported by Andy Bierman).
>     Fixing ietf-yang-push.yang and adding the following before element
>     "datastore-xpath-filter " fixed the example.
>         <yp:datastore
>         xmlns:ds="urn:ietf:params:xml:ns:yang:ietf-
>      datastores">ds:running</yp:datastore>
Alex is going to be making the change to the YANG push datastore drafts shortly (I just talked with him).  I am happy to modify the example (if necessary) based on  the "when" statement incorporated under update-policy-modifiable.  

>     Minor comments and nits:
>     - Abstract s/YANG Push/YANG-Push/ since YANG-Push is the term introduced
> in
>     draft-ietf-netconf-yang-push 
Done

>    Introduction. Last sentence doesn't read well,
>     "...a RESTCONF client via a dynamic subscription..." may need tweaking -
That text has been touched so many times, I am afraid of touching it again.

>     Section 5 s/terminates/terminates,/ 
Done

>     Section 7 3rd bullet. s/this MAY but does
>     not have to be included/this MAY be included/? 
Done

>     Section 7
>     s/kill-subscription-error/delete-subscription-error/ (thanks Rob) 
Done

>     Figure 3,
>     stream-xpath-filter uses prefix "ds" for example-datastore.yang. Since "ds" is
>     used by ietf-datastores.yang, might be better to choose another prefix such
>     as  "ex"
Done.

I will post after a couple days, or after the "when" question closes (whatever comes first).

Eric

>     Regards,
>     Reshad.
> 
> 
>     _______________________________________________
>     yang-doctors mailing list
>     [e-mail]
>     https://www.ietf.org/mailman/listinfo/yang-doctors
>