1. Pages
2. …
3. Open Transport Configuration and Control Home
4. TAPI
5. TAPI Call-in Details and Notes

2021-05-04 TAPI Meeting Notes

All

• Delivery plan (State of RIA and 2.3)
  • Andrea Mazzini TAPI 2.3 RC1 - end of next week
  • Ramon Casellas RIA 1.1 - end of next week
    • How to include the alarm spreadsheet in the delivered TR-547 pdf file: by reference to a separate read-only Excel file. For further analysis.

11 May 2021 TAPI Call: 2 hours

Karthik Sethuraman and Nigel Davis are looking at O-RAN.

Karthik Sethuraman is involved in the O-RAN evaluation of TAPI Topology model. So far mostly admin discussion, e.g. how to adopt/formalize usage of TAPI.

• O-RAN is considering topology not only regarding forwarding. In TAPI the focus is on forwarding and data plane model, while O-RAN is considering also Control Plane aspects.
• At the bottom the technology is L2/VLAN switching, then the client is L3/IP.
• In O-RAN seems missing the definition of an overarching architecture, many teams are developing distinct aspects.
• Noted that O-RAN software is open, while O-RAN standard activity is closed.
• Nigel Davis in Core IM the Control Node and Processing Construct can be reused for Control Plane aspects.
  • Andrea Mazzini , Karthik Sethuraman we need a couple of slides to communicate how these Core IM concepts can be applied.

Andrea Mazzini  presents proposal_tapi_yaml_correction_post_body.docx

Agreed the proposed modification of TapiConnectivity Yaml module.

From:

           - in: "body"
        name: "tapi.connectivity.ConnectivityContext.body-param"
        description: "tapi.connectivity.ConnectivityContext to be added to list"
        required: false
        schema:
          $ref: "#/definitions/tapi.connectivity.ConnectivityContextWrapper"

to

      - in: "body"
        name: "tapi.connectivity.ConnectivityContext.body-param"
        description: "tapi.connectivity.ConnectivityContext to be added to list"
        required: false
        schema:
          $ref: "#/definitions/tapi.connectivity.ConnectivityContext"

Ramon Casellas mentions the name space, it should be:

• "tapi-connectivity:connectivity-service": [

For further check, as in the Yaml modules the name space is present.

Discussion on the added value of TAPI OAS, recalled that - as stated in release notes:

• The TAPI YANG models included in this TAPI release (vx.y) are a normative part of the TAPI SDK. 
• TAPI OAS (OpenAPI Specifications) included in this TAPI release (vx.y) are an informative part of the TAPI SDK.

Ramon Casellas also recalls that the decision to remove "GET on lists" was justified by limits of XML encoding, while with JSON encoding there are no issues.

Andrea Mazzini proposed to move FEC related parameters either to OTU CEP or OTU MEP.

• Agreed the OTU MEP, hence the threshold values can be provisioned at Connectivity Service creation, through CSEP augmentation:
  
  • ConnectivityOamServicePoint - OduOamMepServicePoint - OtuMep

Hing-Kam Lam highlights that ES counter has been deprecated by ITU-T, according to contribution C605 (from Bernd Zeuner to the 3/2014 ITU-T SG15 meeting). The proposal was agreed by Q14/15 and led to the removal of near-end (nES) and far-end ES (fES) in G.874 since its Amendment 1 (8/2015).

• ES at OTUk and OTUkV:

  • G.798 defines MI_pN/F_EBC for OTUk and OTUkV layers. It would be possible to calculate ES but it is rather likely that ES occurs every second and is therefore as meaningless as for ODUk.

Andrea Mazzini briefly presents TAPI2.3AgreedUpdates.docx

The document collects all agreed modifications to UML and RIA as tracked by TAPI call minutes since July 2020.