Re: [Netconf] I-D Action: draft-ietf-netconf-rfc7895bis-05.txt

[Ladislav Lhotka <[e-mail]>]

Juergen Schoenwaelder <[e-mail]> writes:

> On Mon, Mar 05, 2018 at 05:12:09PM +0100, Ladislav Lhotka wrote:
>> On Mon, 2018-03-05 at 17:01 +0100, Juergen Schoenwaelder wrote:
>> > On Mon, Mar 05, 2018 at 04:50:51PM +0100, Ladislav Lhotka wrote:
>> > > > 
>> > > > Note, that the deviation leaf-list is a leafref back to the module 
>> > > > list.  Hence, for the YANG instance data to be valid, any entries in the 
>> > > > deviation list must be included in the modules list.
>> > > 
>> > > Hmm, even this is actually not true - the type should have
>> > > 
>> > >   require-instance true;
>> > > 
>> > > right?
>> > 
>> > Yes, I think this should be there.
>> >  
>> > > If "module" and "deviation" are required to match, the document
>> > > should say so.
>> > 
>> > Note that section 3 says:
>> > 
>> >    o  deviation list: The name of each YANG module with deviation
>> >       statements affecting a given YANG module, in a given datastore
>> >       schema, MUST be identified.
>> > 
>> > I know, this is in the objectives section and not in the YANG
>> > definitions, perhaps we need to make this clear there as well.
>> > Perhaps just change
>> > 
>> > OLD
>> > 
>> >      "List of YANG deviation modules used by this server to modify
>> > 
>> > NEW
>> > 
>> >      "List of all YANG deviation modules used by this server to modify
>> 
>> But still it doesn't say how the deviation list is determined from YLbis data. A
>> server/client implementor may IMO choose any of the three options that I listed.
>>
>
> A server has to list _all_ deviation modules and I thought this takes
> care of things. The reason for having this list is that a client
> interested in say ietf-interfaces can directly find the modules that
> contain deviations affecting ietf-interfaces. Hence a client has to
> trust this list to be complete, i.e., the server has to report _all_
> deviation modules. If you still feel something is missing, perhaps it
> helps if you propose concrete text.

The server can include deviation modules in the YLbis "module" list. Now, the
question is whether this is enough, no matter what the "deviation"
leaf-lists say. It makes no sense to propose any text before we agree on
the answer to this question.

Note that one module can possibly include deviations for multiple target
modules. An implementor may then decide to apply deviations only to some
modules, and this can be effectively done through the "deviation"
leaf-lists.

Lada

>
> /js
>
> -- 
> Juergen Schoenwaelder           Jacobs University Bremen gGmbH
> Phone: [phone]         Campus Ring 1 | 28759 Bremen | Germany
> Fax:   [phone]         <https://www.jacobs-university.de/>

-- 
Ladislav Lhotka
Head, CZ.NIC Labs
PGP Key ID: 0xB8F92B08A9F76C67