Event Survey

• Overview of stability and resiliency tests since Guilin

• Overview of  new upgrade/rollback functionality / Demo

• Presentation of DT's ONAP based "TNAP" portal architecture.
  • Midterm plan to contribute the portal code to the community (not yet decided and scheduled)
• Real Demo!

• Argo CD helps us, doing GitOPs in the "TNAP" project and install ONAP and TNAP components.
• Real Demo!

• Presentation on E2E Network Slicing Overview and Istanbul release highlights
• Demo on the below scenarios
  • NSMF driven TN Slices and E2E Network slice allocation
  • E2E Network slice allocation and reuse with all internal NSSMFs (RAN NSSMF driven TN-FH & TN-MH)
  • E2E Network Slice Reuse using external NSSMFs (Option 2)

• Overview of Open and Disaggregated Transport Networks
• Introduction to ONF T-API YANG models
• Explanation on Nokia contribution to tapi NBI of ODL/TransportPCE:
  1. T-API topology service
  2. T-API connectivity service
  3. T-API notification service
• Demo showing the creation of an OTN 10GB service using T-API interface

1. Conversations needed in order to position the proposed extended optical SDN controller in the Transport Slice use case of ONAP

• Provided high-level overview of Verizon's Opendaylight scaling architecture.
• Discussed the important blocks of the diagram along with the scaling procedures.
• Good questions from the community about callhome feature support and how to move forward. 

1, Promoting more close collaboration between vendors and operator by continuous integration multiple vendor products, continuous testing, continuous optimizing.

2, Promoting Pre-integration and Pre-deployment before production deploy, will help improve the quality and save time.

1, Openstack deploy could be very complex, hard to fully automate deploy and upgrade

2, Not easy to define which vendor should be responsible for the defect, suggesting apply RC testing.

Overview of AAP, and overview of the value proposition

• Cost and time savings
• Efficiency/Simplicity  

Discussions/feedback from Operators and vendors as well as other Communities

Transport-ability and ease of adoption added to the list of values

1) Develop a cross LFN community strategy

2) Assist Telcos with RFP/RFI questions

3) Outreach 

4) Include Capacity testing

Agreement on RC2 Scope and differentiation between RC2 and AAP

Issues still to be resolved on Traceability and Release Mgmt.

Various MEC scenarios

Complexity of networking to address various MEC scenarios

Need for automation

Enhancements in EMCO to address traffic steering scenarios between 5G and MEC applications

Supporting MEC is divided into four phases

Phase 1:  Supporting Local MEC

Phase 2:  Multiple instance of apps across MECs

Phase 3:   Security as a Service for MEC applications.

Phase 4:   Selection of application instance based on 5GFF EDS

Network slicing requires many projects to work together.

Network slicing is realized using EMCO, Nodus, SDEWAN, ExternalDNS, MetalLB, Cert-manager and other CNCF projects.

Session went through the challenges faced in integrating all of these together, problems founds and fixes made by different communities.

More can be done.  2022 plan

• full automation
• Slicing orchestration

GitOps model is popular way of synchronizing resources in target K8s clusters.

Azure Arc, Google Anthos, AWS managed K8s support Gitops model.

EMCO enhancements are discussed to work with GitOps

2022 will support

• Azure ARC
• Fluxv2
• Google Anthos

• Quick walkthrough of collaboration with Akraino and Tungsten Fabric

• Discussed evolution of Network slicing; current and future capabilities.
• Demo of EMCO orchestration of slices & MEC applications

• Automation needs.
• Performance isolation needs
• Security isolation needs
• Type of observability & closed loop actions needed to maintain SLAs even when there are noisy neighbors.