Hi Juergen,

Consider a device supports a dynamic data-store called "ephemeral". Consider 
Yang-library 1.1 is NOT implemented in device.
Consider device have 3 Modules ==> Module1 ,  Module2-state and Module3

Module1 : Supported in <running> and <operational>
Module2-state : Supported only in <operational>
Module3 : Supported only in <ephemeral>

So here assumption is that Client knows he cannot do <edit-data> on 
Module2-state as it has only read-only nodes.
But Client does not know the modules in ephemeral and depends on yang-library 
1.1 to know whether an edit-data can be done on Module3 ?

Which modules are supported in which data-stores is not known to Client without 
Yang-library 1.1 ?

With Regards,
Rohit R Ranade

-----Original Message-----
From: Juergen Schoenwaelder [mailto:[email protected]] 
Sent: 29 May 2018 16:36
To: Rohit R Ranade <[email protected]>
Cc: Robert Wilton <[email protected]>; [email protected]
Subject: Re: [netmod] draft-ietf-netconf-rfc7895bis-06 deviation query

On Tue, May 29, 2018 at 10:44:04AM +0000, Rohit R Ranade wrote:
> Hi Robert,
> 
> The Introduction section has :
> "
> Furthermore, the operational state datastore may support non-configurable 
> YANG modules in addition to
>    the YANG modules supported by conventional configuration datastores.
> "
> I infer that in the new Yang-library structure,  the schema for 
> "conventional" data-stores should not include the non-configurable YANG 
> module. Is my inference correct ?
>

A module that has only config false nodes will not add anything to a 
conventional configuration datastore but it is not an error list such a module. 
For a simple devices, it may be desirable to have just one schema that applies 
to all datastores. There is no requirement to break things apart just because 
there is a module that contributes only an empty set to conventional 
configuration datastores.

Note that the defined term is 'conventional configuration datastore'
and not 'conventional datastore'. Oops, I see that in one place 
draft-ietf-netconf-rfc7895bis-06.txt missing the 'configuration' part of the 
term. That should be fixed.

OLD

   The recommended approach to populate "/modules-state" is to report
   the schema for YANG modules that are configurable via conventional
   datastores and for which config false data nodes are returned via a
   NETCONF <get> operation, or equivalent.

NEW

   The recommended approach to populate "/modules-state" is to report
   the schema for YANG modules that are configurable via conventional
   configuration datastores and for which config false data nodes are
   returned via a NETCONF <get> operation, or equivalent.

Co-authors, if you agree, how do we track this?

/js

-- 
Juergen Schoenwaelder           Jacobs University Bremen gGmbH
Phone: +49 421 200 3587         Campus Ring 1 | 28759 Bremen | Germany
Fax:   +49 421 200 3103         <https://www.jacobs-university.de/>

_______________________________________________
netmod mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/netmod

Reply via email to