Re: [netmod] YANG actions - need to define OK/error or can reuse NETCONF ok/rpc-error ?

[Martin Bjorklund <[e-mail]>]

"Sterne, Jason (Nokia - CA/Ottawa)" <[e-mail]> wrote:
> Thanks Martin.  I should have found that myself when I was looking
> through that section.
> 
> What about the case where the operation invocation did not succeed and
> there are no output parameters defined ?

In NETCONF that will be a <rpc-error>, as per RFC 6241.  The action is
invoked with a normal NETCONF <rpc> message.


/martin



> 
> Rgds,
> Jason
> 
> > -----Original Message-----
> > From: Martin Bjorklund [mailto:[e-mail]]
> > Sent: Thursday, June 7, 2018 5:21 PM
> > To: Sterne, Jason (Nokia - CA/Ottawa) <[e-mail]>
> > Cc: [e-mail]
> > Subject: Re: [netmod] YANG actions - need to define OK/error or can
> > reuse
> > NETCONF ok/rpc-error ?
> > 
> > Hi,
> > 
> > "Sterne, Jason (Nokia - CA/Ottawa)" <[e-mail]> wrote:
> > > Hi all,
> > >
> > > When defining an 'action' in a YANG 1.1 model, and we want the server
> > > to be able to respond with <ok> or some error information, do we need
> > > to define the ok/error info in the 'output' of the action, or can we
> > > define an action without any 'output' statement and have the server
> > > respond using the typical <ok> or <rpc-error> in NETCONF ?
> > 
> > You don't need to define "ok" - see section 7.15.2, the last paragraph
> > of RFC 7950:
> > 
> >    If the action operation invocation succeeded and no output parameters
> >    are returned, the <rpc-reply> contains a single <ok/> element defined
> >    in [RFC6241].  If output parameters are returned, they are encoded as
> >    child elements to the <rpc-reply> element defined in [RFC6241], in
> >    the same order as they are defined within the "output" statement.
> > 
> > 
> > 
> > /martin
> > 
> > > I'm not sure if it is relevant, but when I look at the definition of
> > > the commit rpc in the NETCONF spec, there is no 'output' defined but
> > > clearly a response of <ok> or an rpc-error can be returned by a
> > > server.
> > >
> > > If we don't define the ok/error in the action itself then I suppose
> > > other types of interfaces (RESTCONF) may or may not have other ways to
> > > reply ok/error (at least it won't be defined by the YANG model for the
> > > particular action).
> > >
> > > But it does seem like a waste to go and specify ok/error information
> > > for every action out there if they only need to return ok or error
> > > information that could be carried in the standard rpc-error message.
> > >
> > > Rgds,
> > > Jason
>