[OPSAWG] Soliciting thoughts on draft-ietf-netmod-yang-model-classification

["Adrian Farrel" <[e-mail]>]

Hi,

We previously discussed draft-wu-opsawg-service-model-explained a bit on this
list and also at a face-to-face meeting.

Since then the I-D has moved on to a -03 revision picking up and addressing
comments from the list, private emails, and the authors' gradual understanding
of the field.

We do not think that this I-D is conflicted by, with, or from
draft-ietf-netmod-yang-model-classification (which we understand may be
approaching WG last call in NETMOD).

We believe there are some issues that would benefit from focused discussion in
the OPSAWG.

Is there agreement on our separation of models into customer service models,
service delivery models, and network/device configuration models? We think that
we have the right split of function, but not necessarily the right names for the
terms.

Can we refine the definition of "service"? There was some discussion of this at
the last meeting, and it led us to refine the language a bit, but it would be
helpful to have more feedback.

Is this stuff "obvious" or is there value in publishing an RFC? Our experience
working in L3SM has been that there is a tension with the protocol working
groups (particularly BESS) that arises from the equipment vendors being closer
to how services are delivered within a network (i.e., by configuring a set of
devices in the network) than they are to an understanding of how an operator
presents a service for consumption by their customer. Our intention is to make
that distinction clearer.

Have we got the correlation with the MEF's LSO right? We have heard a
conflicting view to what we have written, but we have also heard supporting
views. Should we drop all discussion of the MEF's model as it is out of scope
for the IETF?

Thanks for any opinions.

Adrian