Minutes for the Network Modelling Language WG session #1 (2008-06-02)

Paola Grosso introduces the Agenda, and repeats the goals that have been set for this WG in the charter.
Work on the deliverables is ongoing. Deliverable 1 is almost finished, and is currently awaiting final input from the Martin Swany (as representative of the perfSONAR group). When he is done, it can be put into the OGF process of making this an official document.

The contents of Deliverable 2 (the layer-agnostic schema) is the focus of the meetings at this OGF.

Then there is presentation by Freek Dijkstra (UvA) about lessons learned in modelling multi-layer inter-domain optical networks. At the end of his presentation, he was asked for the most important points:

- There needs to be a distinction between models and syntax (so that it is possible to have different syntaxes based on the same model)
- What must be described will change over time, it is therefore necessary to be able to have a model that is extensible. And Freek claims that this is possible to do without introducing incompatibilities.
- It is necessary that the syntax can integrate with other kinds of resource descriptions, so that resource schedulers can take different types of resources into account.

Martin Swany presents an overview of the (new) WGs (NM-WG, NML-WG, NMC-WG, DMN-WG (Dynamic Management of Network Resources, (name still under debate)?)) and how these relate to each other.
He also presents the concept of NURNs (Network URNs). Later on Yuri remarks that the use of URNs is not currently possible within the OGF. This needs to be researched further.

Finally he presents a basic set of elements of the perfSONAR/NM-WG/Dice/... schema. This will be used as one of the basis for tomorrows discussions.

Due to time constraints we cut the discussion short, and move over to Yuri Demchenko's presentation. He presents research done to identify attributes for authorization in network resource provisioning. This is a work in progress to create a policy framework for getting network access. He has been able to use some of the currently existing network description syntaxes as a basis for describing policy in XACML.

Richard Hughes-Jones remarks that it is very important to be able to test a connection once it has been provisioned. And this is a message that Yuri will bring back to the Phosphorous group.