Skip to end of metadata
Go to start of metadata


12 November 2018


  • Lyndon Ong

  • Karthik Sethuraman

  • Andrea Mazzini
  • Bernd Zeuner
  • Kam Lam
  • Yossi (Ceragon)
  • Thorsten Heinze
  • Italo Busi


Discussion Items

 ONF PRLyndon
 ONF Connect TalksLyndon
  • ONF Connect presentations also approved on TAPI, CIM and WT work
 External Activities 
  • OIF - planning stages for possible TAPI certification program
    • 8 carriers expressed interest in participating in program development
    • Certification around TAPI 2.1
    • More discussion and work planned into 1Q19
  • OIF Liaison to ONF approved
    • carries specs used in the 2018 OIF demonstration
  • ITU (Kam)
    • available on liaison wiki page: Liaisons
    • some material on photonic model - related to TAPI work
    • other material aimed at CIM modeling work
  • MEF (Karthik, Andrea)
    • TAPI-based specs (NRM OAM, NRM L1) into next stage of comments, some discussion about alignment of core models
    • Wide consensus on the need for tooling for UML, e.g, to support OCL - next steps tbd but could be additional MEF support for expanding tooling
      • part of wider interest in tooling including for other data modeling, e.g., TOSCA
      • Q on OpenAPI - MEF using a variety of tools depending on the project, EAGLE (same as used for TAPI 2.1) used in NRM work, some groups do not use YANG
  • Other
 Internal Activities 
  • TAPI
    • Karthik working on RI based on interest from ONF ODTN group, could be used as emulator - any help welcome
  • OTIM
    • continuing coordination meetings with other SDOs, e.g., IEEE, ITU-T
  • WT
    • next PoC in 2 weeks in Munich, variety of use cases, expecting 20 participants including 3 operators and 17 vendors, each vendor will support at least 2 use cases
  • DMIP
    • to be tested in PoC depending on level of implementation
 Future planning 
  • Next OTCC call scheduled for 11/26
    • no TAPI on Tuesday this week, and probably next week - use Thursday timeslot instead - Karthik will send email on this - may want to consolidate call times in future (some support for this)
  • F2F week of 12/3-7
    • note overlap with ONF Connect sessions and other ONF activities, esp.
      • Monday 9:30-1:30 ODTN Use Case group plans to meet in Menlo Park ON.Labs office - possible joint session in the afternoon? invite them to join F2F at least
      • Wednesday 2pm on is the ODTN session
      • Thursday 4:30 is the ODTN Reference Design review
      • ONF Connect talks as discussed above
    • current plan: Agenda plan

Thorsten - Q. on technology-specific models for the operator - issue brought up earlier - need for model "landscape" using core model to consolidate technology-specific models, to support operator multi-technology network control - possible sources could be IETF (but concerns with need for proprietary extensions), OpenConfig (no core model), ONF (good core model but need more technology-specific extensions)

OpenConfig seen to have an advantage due to the broad support of different technology-specific models - how can we extend ONF work? Propose guidelines for making technology-specific amendments to the Core Model, a cookbook that can be used by other groups/operators. Could initially cause some redundant developments, but ONF could follow up by some type of validation that guidelines are fulfilled, listing on ONF website, then identify what models are being supported across multiple operators/vendors, track the acceptance in the industry - goal to accelerate extension of the core model across technologies. Concern that if we are not able to do this, CIM work will become irrelevant.

What is needed beyond what is in TR512.7 and other documents? Should the models be defined in associated SDOs first? With TAPI, have worked closely to ensure consistency. What other technologies are needed? Thorsten - technically easy to add, but industry does not know this - needs to be made known to others that you don't need to understand all of the core model. Reality is that ONF is not seen as the place to solve modeling problems, compared to OpenConfig which identifies dozens of models on their website, even if each model is very limited. ONF appears too academic and not a place for solutions at this time. TAPI is known as a solution, also microwave model, but lots of areas not addressed (e.g., synchronization). Note operators may have different requirements on their models also, some may want only very simple models, or different needs depending on the application.

Could maybe start with OpenConfig models, integrate them with core model? Further discussion needed, item for OIMT as well. Kam and Thorsten to coordinate, possibly next Thursday (11/22) OIMT-E.

Action Items