Skip to main content

Liaison statement
Liaison from ONF architecture WG

Additional information about IETF liaison relationships is available on the IETF webpage and the Internet Architecture Board liaison webpage.
State Posted
Submitted Date 2014-10-27
From Group ONF-ARCH-WG
From Contact Dave Hood
To Group IETF
To Contacts The IETF Chair <chair@ietf.org>
Cc The IESG <iesg@ietf.org>
brian@innovationslab.net
ted.lemon@nominum.com
bclaise@cisco.com
joelja@bogus.com
akatlas@gmail.com
adrian@olddog.co.uk
granville@inf.ufrgs.br
olivier.festor@inria.fr
Response Contact Liaisons@opennetworking.org
Purpose For information
Attachments Liaison from ONF architecture WG
IETF_Liaison 20141023 cover letter
Body
Dear colleagues,

A critical aspect in the successful realization of the SDN vision is a
comprehensive modeling description of network resources (functionality,
capabilities, flexibility) that enables the necessary level of control/network
programmability. This is essential, given the need to enable provision of new
services and applications that may rely upon support from a wide range of
networking domains involving heterogeneous technologies. Further, the
boundaries of networking domains are tending to blur, in terms of definition of
access, metro, core and what technologies are deployed in which networking
domain. Without such an underlying common information model, it is anticipated
that SDN deployments will encounter the same issues that have previously been
faced by management system deployments in attempting to achieve global
information networking.

Thus, a key enabler to successful worldwide adoption and deployment of SDN is,
in fact, a common information model of network resources. However, as in
software development everywhere, in order to assure successful execution in a
distributed world, a common information model cannot be a monolithic entity.
Rather, it must be structured in a modular way, so as to allow for independent
specification development efforts, while providing a framework that assures a
coherent whole. Recognizing this, the ONF is developing a supporting process
for Common Information Model development, described in onf2014.431 (attached in
Annex 1).

The Common Information Model is structured in terms of a core model fragment,
forwarding technology specific model fragments, and application specific model
fragments (not precluding other types of fragments). Artefacts in the core
model fragment are usable by groups/organizations working on forwarding
technology or application specific modeling activities. It is recognized that
the core model fragment itself may evolve; e.g., due to discovery of additional
common artefacts from evolving work across domain specific modeling activities.
Currently, there are activities ongoing in ONF on the core model (Architecture
WG), L0-L2 transport (Optical transport WG), mobility (Wireless & Mobile WG),
and NorthBound Interface applications (NBI WG). One of the major strengths of
the adopted process is that there is no need for all domain specific modeling
activities to reside in a single group within a given organization, or for that
matter, in any single organization.

In addition to information about ONF’s Common Information Model development, we
also enclose draft document onf2014.491, containing the UML modeling guidelines
that are used by the modeling teams in the ONF. The conventions ensure
consistency across all model development, including but not limited to naming,
data types, graphic representation and UML artefacts.

We also enclose draft document onf2014.505 containing the guidelines for using
the open source tool Papyrus for development of UML information models. These
guidelines have been adopted by all of ONF; i.e., they are not specific to any
working group, technology or management protocol.

Finally, we also enclose a snapshot of the information model, as it stands
today.

All of these documents represent work in progress, to which the usual
disclaimers apply. Given the intention that the core information model be
valuable across the industry, not just locally within ONF, we encourage IETF to
adopt any of the tools, concepts, or  ways of working in this work. We also
invite feedback on any of these documents, and we welcome opportunities for
collaboration.

Sincerely,
Fabian Schneider, Chair ONF Architecture WG
Dave Hood, Vice-Chair ONF Architecture WG
Kam Lam, Vice-Chair ONF Architecture WG