Skip to end of metadata
Go to start of metadata

Date

Attendees

Goals

  • General update
  • TAPI Hybrid - Equipment and OAM augments to TAPI 2.1.3
    • Feedbacks from TIP MUST
    • Photonic OAM vs. Inventory - RIA 1.2 (and 2.1) may make PM metrics as optional in Inventory UCs
  • Path Computation UCs
  • Model of the internal connectivity matrix of a ROADM
  • OAM and Protobuf/gNMI encoding
  • Continue the investigation on TAPI-IETF navigation
  • Access to the photonic plug
  • AOB

Agreed Items & Priority

Discussion items

5 minsAdministrative

All



TAPI v2.4.1  has been finally delivered - thanks to all the Team!

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
  • Model of the internal connectivity matrix of a ROADM
  • OAM and Protobuf/gNMI encoding
  • Continue the investigation on TAPI-IETF navigation
  • Access to the photonic plug
5 mins

General update

Andrea Mazzini recalls that TAPI 2.4.1 and RIA 2.1 have been officially delivered.

  • This 2.4.1 version is candidate for adoption and implementation.
20 mins

TAPI Hybrid - Equipment and OAM augments to TAPI 2.1.3

  • Feedbacks from TIP MUST
All

Andrea Mazzini shows the Arturo's email with the TIP MUST position, copied below:

The MUST group has been informed about the proposal raised by Telefonica to the ONF TAPI group to extend the TAPI v2.1.3 branch to cover the following uses cases:

      • OAM Hybrid: OAM Job (2.4) on CEP (2.1.3)
        • OAM Job exists in TAPI 2.1.3 but it was related to MEP and MIP only.
      • physical-route-list (2.4) on tapi-connection (2.1.3)
      • access-port-supports-sip (2.4) on SIP (2.1.3)

We understand that the proposal includes some augments which will not introduce backward compatibility issues with current v2.1.3 implementations.

The group should still analyze the proposal in more detail and agree upon its adoption within MUST Technical Requirements for Optical/O-OLS SDN Controller NBI. Furthermore, the group does not commit to fostering the adoption of this new release rather than v2.4 and v2.5 releases. However, the group acknowledges the technical validity of the proposal and its purpose to accelerate certain use cases driven by business needs. 

Esther adds that it is not the intention of TIP MUST to block this initiative.

  • ONF OTCC will follow up with TIP MUST on the subject.
10 mins

TAPI Hybrid - Equipment and OAM augments to TAPI 2.1.3

    • Photonic OAM vs. Inventory
All

Confirmed that it is necessary to distinguish between different natures of the modeled items:

    1. Configuration parameters
    2. Stable resource state parameters (e.g. configured mapping types, thresholds, etc.) in general the resource state parameters driven (mainly) by the configuration parameters
    3. Not stable / extremely unstable resource state parameters (e.g. operational state, PM Metrics, etc.) in general the resource state parameters driven (mainly) by network events
  • Each parameter type requires different access methods.
  • Andrea Mazzini this approach shall be reflected in the proper sections of the RIA.
    • Future evolution of the model shall improve the decoupling of parameters having different natures, to facilitate the differentiation of the access methods.
80 mins

Path Computation UCs

All

Andrea Mazzini recalls the existing documentation regarding Path Computation:

  1. otcc2020.ND.010_TAPI-ConnectivityServiceComputationService.pptx

  2. otcc2021.RC.001_TAPI-PathComputation.pptx

  3. RIA 2.1 UCs:
    • Use case 12a: Path Computation
    • Use case 12b: Simultaneous pre-calculation of two disjoint paths
    • Use case 12c: Multiple simultaneous path computation (Bulk request processing)

Ramon Casellas recalls that one of the possible original sources of the Path Computation model is IETF PCEP (RFC 5440)

  • The Explicit Route Object (RFC 3209) is an ordered list of numbered or unnumbered interfaces - something similar to a list of TAPI NEPs or CEPs if considering the (G/MPLS) label.
  • Andrea Mazzini in general we need to identify and review the IETF references, for a better understanding of the original requirements.
  • Agreed that the TAPI model shall allow the specification of label/timeslot/wavelength constraints, besides Link/NEP.
  • Discussion on Layer Protocol Name and Qualifier of the Path Computation Service.
    • Preliminary agreement that the layer attributes shall indicate the preferred or unique layer of the Link/NEP composing the resulting Paths, e.g.
      • ODU4 LPQ shall drive the creation of Path(s) composed by Links supported by ODU4 Connections (or equivalent NEPs), see top left Link in the figure:

Current model foresees that the PSEP (Path Service End Point) shall refer to a SIP.

  • Preliminary agreement that the PSEP can refer either to a SIP or to a NEP, to improve provisioning flexibility.
  • Ramon Casellas suggest to keep the currently defined referece to the SIP, for the case of opaque view of the topology.

Agreed to postpone the "top down" feature, where the Server Controller is in charge to complete the connectivity of a virtual network,

  • e.g. if an ODU4 trail is missing to complete a Path, then it is automatically created by the Server Controller. In fact, Esther underlines that the original Use Case is simpler.

Andrea Mazzini presents the PathSet model, originally proposed by Nigel Davis in otcc2020.ND.010_TAPI-ConnectivityServiceComputationService.pptx.

  • The PathSet and PathSetConstraint are defined to implement Use case 12b: Simultaneous pre-calculation of two disjoint paths

Ramon Casellas clarifies that the Use case 12c: Multiple simultaneous path computation (Bulk request processing) may need another grouping construct, e.g. a PathComputationServiceGroup to allow the bulk provisioning of more Path Services, allowing the Server to perform a global concurrent optimization.

  • For example, the disjoint provisioning of Path Computation Services may allocate network resources in a locally optimized way, making increasingly difficult/costly the computation of further Path Services.