25 June 2018 OTCC TST Call
OIF Demo Update
|Face-to-face Meeting||Lyndon, Kam, Nigel, Karthik|
Met June 11-15 in Ottawa
WT - TR-532 - elaborated issues list with 37 issues from PoCs, working through the list and proposing solutions - available on the ONF wiki for comments, to be discussed in WT call - planning to document as v1.1, backward compatible - for 5th PoC
DMIP - TR-545 - many comments on first review, completed 2 weeks ago and another review initiated within DMIP group. 2nd review at DMIP level being held with deadline end of this week. After resolution, this would then go through OTCC review. Aim at middle of July for completion.
Ethernet PHY model - updated with WT testing results, removed some sections (some now covered in TR-512.6 physical model). Will propose as candidate for 5th PoC for management of tributary ports.
Possible interest in DMIP from other groups? tbd (maybe ODTN?)
TAPI - 80% complete on OAM model, still working on agreement on the photonic model - also need UML-to-YANG tool improvement, for YANG-to-Swagger plan to use MEF/ODL tool, but also now some possibility that the pyang tool will be updated, may need to select one in future.
|Other Business||Thorsten||Would like to attach technology-specific definitions to the models, but notes many people point to OpenConfig modeling - but also much missing, no core information model equivalent, less advanced equipment modeling - ONF CIM and technology-specific modeling is not that well known - can we make this work better known to people in the industry? Not really visible, e.g,. on the ONF website and not being publicized/socialized. LTP and other concepts are quite powerful and should be useful for others as well. Example - could be applied to CPRI. Need more operator endorsement? OpenConfig is use case driven, but not considering the basic modeling as much. Italo notes concern with duplication of work leading to more that vendors need to support, Thorsten believes there is not actually that much duplication because the focus of different efforts are not the same. Thorsten notes that the ability to learn new capabilities is an important part of the model. Nigel notes that we point to ITU-T whereas OpenConfig tends to duplicate work already done in ITu-T. Work should go back to the group with responsibility for that technology.|