1. Pages
2. …
3. OIMT
4. Core Model
5. OIMT minutes
6. 2018 OIMT Minutes

2018-02-01 OIMT Meeting Notes

TOSCA and Association constrain

Nigel noted that Chris Lauwers is using Papyrus to capture the TOSCA metamodel and that he may want to leverage our work.

Action: Nigel: Contact Chris Lauwers regarding Tosca work in this area.

Topology example usage (not covered in the meeting)

• Action item - Nigel: Provide some examples.
  • • • May look at the IETF cost parameters
      • Start with a simple example: e.g., the cost of an unprotected link
      • https://tools.ietf.org/html/rfc7285

Control Model

• Control component is a management agent or perhaps a control agent?? Chris will explore for alternative names.
• ControlSystemView = ControlDomain. It is just a collection of the things that can be accessed.

Software (TR-512.12)

Action item - Nigel: Need to enter model into Papyrus from slide pack (from last face to face).

Wireless Transport discussion

• Model backward compatibility
  • Martin: Challenge... we have two levels. UML and Yang.
  • Nigel: we need to cover both
  • Martin: In UML we have all the mechanisms.
  • Nigel: Semantic boundary should not be violate.
  • Martin suggested that Netconf.
    • Please see Guidelines for Authors and Reviewers of YANG Data Model Documents
      • Section 4.6: Lifecycle Management
    • Please see YANG 1.1 Section 7.21.2: The "status" Statement
    • Please see YANG 1.0 Section 7.19.2: The status Statement
  • Nigel noted tooling and translation.
  • Martin indicated that the tooling was the key consideration
  • Nigel noted that we should have a diff
  • Martin noted that 1.2 --> Yang there was a list of LTPs in an NE. With the new model, there is a root container and then there are all the lists on the same level that only work with references.
    • The tool is maturing. The tool produces an output that is more different than the differences just that caused by the model.
    • UML - Yang guidelines also are maturing
  • Bernd asked whether I mean schema tree and data tree consideration.
  • In some cases there can be different schema structure that produce the same instance structure (i.e. there is no run time incompatibility).
  • Tooling change may cause major changes in the run time schema or minor changes in the run time schema.
    • semantics the same, syntax is of a minor difference (e.g. a small tag extension)
    • Different revision statement in Yang
    • Discussed Yang versioning.
  • We need to abide by Yang versioning
    • The tooling should deal with version changes by assisting in exposing changes and potentially coding some rules and recommend whether there is a version update
    • We should recommend tools
  • Potential changes in UML may change Yang. Are there changes in UML that cause major changes in Yang that we should avoid
    • Suggested that we should not pollute UML with Yang restrictions. Hence we do not think that there should be any changes in the UML guidelines here or the approach to UML construction.