On Tue, Oct 31, 2017 at 10:22 AM, Randy Presuhn <
[email protected]> wrote:

> Hi -
>
> On 10/31/2017 10:14 AM, Andy Bierman wrote:
> ...
>
>> The system side effects are irrelevant, but both the same for rpc and
>> action.
>>
>
> Knowing what the system side effects are is *ESSENTIAL* if these
> things are to be of any use operationally.
>
> The only issues relevant to YANG are:
>>     - datastore used to evaluate action ancestor nodes
>>     - datastore used to evaluate input parameter leafref, must, when
>>     - datastore used to evaluate output parameter leafref, must, when
>>
>>
>> These properties cannot be modified by description-stmt.
>> Not now, not ever. Not for rpc. Not for action.
>>
>> It is irrelevant to NMDA whether the rpc or action modifies a datastore,
>> starts a playlist on a jukebox, or makes some toast.
>>
>
> But from the an operational or interoperabilty perspective,
> that information is crucial.
>
>
I am not saying that the description-stmt is not needed.
Of course I need to know if the system will start a jukebox,
make toast, or delete all config in all datastores owned by user "fred".

The old text about resolving YANG for rpc and action is outdated.
Instead of <running> + config=false floating in space, it is all
well-defined
to be in the <operational> datastore now. That is all NMDA needs to convey.




> Randy
>
>
Andy


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

Reply via email to