Context Document ---------------- Finally published! GFD.165 Syntax and Identifiers ---------------------- Freek does not have time for Syntax specs right now. It is also unclear whether this is useful right now. The identifiers part is work in progress. An RFC to request the urn:ogf namespace is being prepared. Network/Topology/Domain ----------------------- Jeroen and Inder present their proposed definitions. Network will not be defined because it has too many different meanings. NetworkDomain can be nested Undecided whether Topology can be nested Jerry Sobieski: We should make a distinction between resources under the same policy umbrella and resources managed by the same mechanism. Shared mechanism = NetworkDomain Shared policy = PolicyDomain Topology, NetworkDomain and PolicyDomain are all examples of Group instances The definitions of Topology, NetworkDomain and PolicyDomain will be changed with Network Element -> Network Object. Layer/Adaptation ---------------- Freek and Jeroen present their proposed definitions. Layer references the ITU-T concept of "Characteristic Information", because it is the best way to generically characterise layers. AdaptationType for the generic type of Adaptations AdaptationService describes a capability of performing an Adaptation (instance of a Service!) Adaptation is "Actual data transport function where data of one port is embedded in the data of another port." There is a discussion about the Layer concept and the CI term. In the end we agree that CI is the best generic term to characterise Layers. Layers don't necessarily have to be collections of Ports. Layers exist without the Ports existing. * Freek and Jeroen will prepare a new proposal. Adaptation proves to be a difficult concept. Different terms are discussed including encapsulation, embedding, and others. In the end there is rough consensus for the current terminology. But it is clear that the concept will need to be clearly described in the schema, including examples. * Jerry and Freek will work on this. Services/Capabilities --------------------- Martin presents his proposed definition of Service. * Martin will investigate whether the Service concept is related to the GLUE schema concept of a Service. NSI Input --------- John Vollbrecht presents some issues that NSI is struggling with regarding topology representation, ownership and pathfinding through these kinds of networks. It is observed that a topology on the Service Plane maps to the NetworkDomain concept (including recursion), and the topology on the Data Plane maps to the Topology concept. The NML is confident that for now the NML terminology provides the NSI with enough handles to map their terminology. We will keep up the interaction to see how this progresses. Second Session ============== All the issues discussed in the previous session are put up for a vote. There is a strong discussion on the concept of Topology. In the end, we end up with 4 different options that are voted on. 1. Topology: A set of Network Elements 2. Topology: A set of Network Elements and the links connecting them <==== rough consensus 3. Topology: A set of connected Network Elements 4. Further discussion There is consensus that the NetworkDomain and PolicyDomain concepts will be included in the schema as examples of Groups, with the definitions as described earlier. * The Layer concept will be refined by Jeroen and Freek. There is consensus on the AdaptationType/AdaptationService/Adaptation definitions as presented by Freek and Jeroen. There is a long discussion on how to handle the uni-directionality of adaptations. The terminology is also difficult, there are some different proposals that are then put to a vote. 1. AdaptationSource / AdaptationSink 2. Adaptation / De-adaptation <==== rough consensus 3. Adaptation / InverseAdaptation * Jerry and Freek will clarify this further in the schema doc. There is consensus on the Service definition as presented by Martin. The syntax representation issue will be deferred for now. Summary of next actions: * Freek and Jeroen will prepare a new proposal on Layer definition * Jerry and Freek will further clarify Adaptations in the schema doc * Martin will investigate whether the Service concept is related to the GLUE schema concept of a Service * Inder and Jeroen will add the agreed definition of Topology, NetworkDomain and PolicyDomain to the schema doc * Freek and Jeroen will add the agreed definitions of AdaptationType, AdaptationService, Adaptation to the schema doc * Martin will add the agreed defintion of Service to the schema doc * Chin and Martin will prepare definitions of Link/Path/Segment * Jerry, Pascal/Sebastien and Freek will prepare definitions of SwitchMatrix and TransferFunction Service Example Volunteers (from OGF27): * Adaptation Service - Jeroen * Switching Matrix Service - Jeroen * Segment Concatenation Service - John * Multicast Service - Petr * Label Conversion Service - Freek * Data Transport Service - Freek * Measurement Point Service * Virtualization Service * Lookup Service - Gigi * Path Finding Service