Re: [netmod] Action and RPC statements

[Robert Wilton <[e-mail]>]

On 31/10/2017 15:35, Kent Watsen wrote:
> Hi Robert,
>
>> 6.2 Invocation of RPC Operations
>>
>> This section updates section 7.14 of RFC 7950.
>>
>> RPCs MAY be defined as affecting the contents of a specific datastore,
>> any configuration datastore (e.g., <edit-config>), or any datastore
>> (e.g., <get-data>).  The RPC definition specifies how the RPC input
>> data is interpreted by the server.
> s/MAY/may/? - is this draft providing for this, or should it come from
> e.g., RFC 7950?
I think that I prefer MAY because we are stating what RPCs are allowed 
to do, but don't feel strongly on this ...

>
>> RPCs definitions that do not explicitly state an affected
>> datastore(s) modify the general operational state of the server.
>> Hence, if any RPC input data relates to data node instances then
>> those would generally resolve to data node instances in the
>> <operational> data tree.
> I reordered first sentence, and added a comma to the second:
>
> RPCs modify the general operational state of the server, unless
> explicitly defined to affect other datastores. Hence, if any RPC
> input data relates to data node instances, then those would
> generally resolve to data node instances in the <operational>
> data tree.
Also OK with me.


>
>
>> 6.3 Invocation of Actions
>>
>> This section updates section 7.15 of RFC 7950.
>>
>> In YANG data models, the "action" statement may appear under "config
>> true" and "config false" schema nodes.  While instances of both
>> schema nodes may appear in <operational>, instances of "config true"
>> schema nodes may also appear in other datastores.
> okay
>
>> Actions are always invoked on a data node instance that exist in the
>> <operational> data tree.  The behavior defined by an action statement
>> is generally expected to affect the operational state of the server
>> rather than directly modifying the contents of any configuration
>> datastore.
> fixed plurality issue in first sentence, and removed wishy-washy
> language from the rest:
>
> Actions are always invoked on a data node instance that exists in the
> <operational> data tree.  Action statements affect the operational
> state of the server.
This wishy-washy language is the tricky bit:

i) Actions are allowed to have any operational impact on the device 
(just as RPCs are), their behavior is not constrained in any way.

ii) However, as far as I can see, it doesn't make sense for an action to 
directly affect the contents of any configuration datastore, that should 
be done via a purpose built rpc (like edit-config).

My text was aimed to allow (i) but not encourage (ii)  ...  If your text 
is sufficient for this then it is fine with me - I agree that it is 
simpler, and simpler is preferable.

Thanks,
Rob


>
>
> What do you think?
>
> Kent // contributor
>
>
>
>