Re: [i2rs] I2RS Interim Meeting - June 1, 2016 - 10:00am - 11:00am - Topic: Ephemeral State Requirements

[Jeffrey Haas <[e-mail]>]

On Wed, Jun 01, 2016 at 05:08:16AM -0700, Andy Bierman wrote:
> On Wed, Jun 1, 2016 at 2:10 AM, Juergen Schoenwaelder <
> [e-mail]> wrote:
> > On Tue, May 31, 2016 at 03:36:40PM -0400, Jeffrey Haas wrote:
> > > I agree that your observation covers the general intent.  The
> > requirements
> > > language above is attempting to be a bit too specific for a given
> > > implementation detail, namely instantiation of ephemeral behaviors in a
> > data
> > > tree.
> > >
> > > Could you please supply alternative text?
> >
> > It is difficult since I am not sure what the original intention
> > was. There is augmentation at the schema tree level and then there is
> > augmentation at the data tree level. My understanding (which may be
> > wrong) is that I2RS primarily looks at data tree level augmentations
> > of operational state data and not so much at configuration datastore
> > data. Part of the openconfig inspired discussion is to do away with
> > the /foo and /foo-state division and that has an impact how the schema
> > trees are constructed.
> >
> >
> This REQ-05 doesn't give any reason why data nodes need to be tagged as
> ephemeral
> in the schema tree, or what it means to be tagged as ephemeral.
> 
> My intention for suggesting an "i2rs:ephemeral" YANG extension was to
> identify
> the I2RS conformance requirements for a server claiming to implement I2RS
> for a specific YANG module.
> 
> A node tagged as ephemeral MUST be accessible in the ephemeral datastore
> using I2RS if the module (or feature) is supported. An untagged node MAY be
> accessible.

This covers the intent properly.

Your suggestion of a mechanism in yang is probably reasonable, but as Jürgen
would say, this is all about the requirements.  So, could you suggest text
covering the requirement?

> An I2RS data model could define read-only nodes.  These nodes could
> be considered ephemeral operational state.  Should these nodes live in
> this datastore or an "operational" datastore?  I don't know.
> That's why we probably need your draft.

If the implementation detail for ephemeral state is that it is part of a
datastore, then having the operational state in the ephemeral datastore
seems to make sense.  However, I find it more likely that we'll need a
mechanism to get the "union view after precedence" (i.e. panes of glass
model) for a conceptional datastore that covers state from local and
ephemeral.

But again, that's a protocol detail which we're trying to avoid in the
requirements.

If you think it belongs in the strawman or another document, we'd love text.

-- Jeff