Skip to end of metadata
Go to start of metadata

Date

Attendees

Goals

  • General update
  • Continue the Path Computation UCs analysis
  • Continue the discussion of adding OMS CEP spec to version 2.1.3
  • Model of the internal connectivity matrix of a ROADM
  • AOB

Agreed Items & Priority

Discussion items

5 minsAdministrative

All



See TAPI Call-in Details and Notes for the RIA Review call coordinates.

Warning: in these weeks the RIA editors are changing or deleting some of these sessions.


TAPI weeky call

Preliminary agenda:

  • Continue the Path Computation UCs analysis
  • Continue the discussion of adding OMS CEP spec to version 2.1.3
  • Model of the internal connectivity matrix of a ROADM
100 mins

Path Computation UCs

All

Andrea Mazzini shows slides 176...191 from updated otcc2022.AM.001_TAPI_RIASlides.pptx

  • Two examples:

1) Provisioning of Path Service, followed by the provisioning of Connectivity Service using the created Path Set as constraint:

2) Provisioning of Connectivity Service with contextual creation of orphan Paths, that can be later adopted by a Path Service:

  • Scenarios have been preliminary agreed.
  • Discussions:
    • Nigel Davisconsidering a given Path Constraint Profile, used as reference for different Connectivity Services will lead to same Paths? Question is related to the abstraction level of Paths.
    • Esther suggests to add photonic specific constraints like the (minimize) usage of 3R and transceiver modes. This to use Path Computation to explore the forwarding capabilities of photonic networks.
      • The Path can be opaque from Node to Node.
    • Esther: evaluate the ordering of Path requests. Identified two possible behaviors:
      • Concurrent global optimization, i.e. the Paths are computed as a set.
      • Sequential computation, with a prioritization of Paths concerning the allocation of resources, e.g. higher priority → better resources, etc.
      • We may agree to add this distinction to the model.
    • Ramon Casellas in case 3 Paths are requested but Server Controller can successfully compute only 2 Paths, is the overall operation acceptable?
      • Andrea Mazzini it may depend on the type of application consuming the computed Paths:
        • in case of resiliency, is clearly not acceptable to get only one Path instead of main and spare ones.
        • In case of planning / exploration, partial replies seem more acceptable.
    • Esther: from operational perspective is really important the identification of the reason for failure, to allow the Client Controller to tune accordingly the further requests.
      • Nigel Davis in some cases the Server Controller may not be able to identify the reason for failure.
  • Next step: produce the draft of the Use Cases, starting from the ones already present in TR-547 RIA.
20 mins

Proposal to add OMS CEP spec to version 2.1.3

Olivier Renais

Discussion on the TAPI 2.1.x development stream, recently reopened to fulfil Telefonica requirements (first time discussed on 2023-03-28 TAPI Meeting notes - TAPI Hybrid).

  • Last week Olivier proposed another enhancement candidate for TAPI 2.1.4
  • Clarified that the current plan does not foresee the backporting of optical impairments model in 2.1.4
  • Olivier: in Transport PCE it is still under evaluation which TAPI version to focus on, there are conflicting aspects.
    • Transport PCE uses TAPI to perform the tunnel request. It is foreseen a huge investment in the code development, hence selecting the TAPI version is critical.
    • One vendor (Ciena) will provide TAPI 2.4.1 implementation in a short time, other vendors seem not.
    • Adding more backporting to 2.1.4 may demotivate the adoption of 2.4.1
    • On the other side, Olivier thinks that the experience in implementing backported features may facilitate future adoption of 2.4.1
  • Andrea Mazzini recalls that the main target of TAPI team is TAPI adoption.
    • On one side, in the past years all efforts have been concentrated in 2.4.1, cleaning the model pattern and solving a number of issues found in previous versions.
    • On the other side, there is the risk to lose the momentum if 2.4.1 is not implemented in a reasonable time.
  • Andrea Mazzini suggests to evaluate more in detail which part of 2.4.1 optical impairments model could be candidate for backporting in 2.1.4