On Fri, Dec 02, 2016 at 01:44:34PM -0800, Andy Bierman wrote:
>
> I have concerns about impact of this work on all YANG-based protocols.
> I have asked several times "how do you decide which servers need to
> implement the intended and applied datastores?" and never got an answer.
>
I do not think people proposed to make the implementation of intended
and applied datastores mandatory to implement. What this proposal is
trying to achieve is to provide an architectural framework that allows
to provide access to additional datastores without having to write
data models in certain ways. Keeping the specific datastore semantics
out of data models (and thereby simplifying data models) is what I see
as the main value of this work.
> I think an Applicability Statement is needed for this work because some
> systems
> converge so fast that the difference between intended and applied
> (for real edits, forget cards not plugged in) will be insignificant.
The current I-D says in section 6.1:
o Support for <intended>, <applied>, and <operational-state> should
be optional to implement.
/js
PS: Do we have to run the discussion via cross-posts on two lists or
can the chairs agree to just use one list, assuming that those who
care can manage to follow this one list?
--
Juergen Schoenwaelder Jacobs University Bremen gGmbH
Phone: +49 421 200 3587 Campus Ring 1 | 28759 Bremen | Germany
Fax: +49 421 200 3103 <http://www.jacobs-university.de/>
_______________________________________________
netmod mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/netmod
