Thierry Evanno

Jonne Soininen

Nokia Local Management Welcome

Nokia Open Source Initiatives

Erwan Duval

Barbara Orlandi

• SDM Expert - Solving data migration challenges
• 5G NetWork Slicing

Each intern will do a 10 minute lightning talk about their project, followed by 3 minutes of questions.

1. Alisha Aneja
2. Girish Lingappa
3. Amir Mohamad
4. Adarsh Pal Singh

Testing pain point and requirement discussion what attendees have. Some topics:

• OpenStack services in containers (apex)
• Functest skips features testing. Should always test
• OpenStack Fenix gerrit patch set to trigger Apex snapshot deployment and testing on top of it

Doctor - Rolling Maintenance and Enhanced Fault Management Demo

I) Doctor maintenance use case demo using new OpenStack Fenix. ETSI defined, OPNFV Doctor refined, Implemented in OpenStack.

II) Doctor enhanced fault management demo. Now many times faster.

Doctor - OpenStack Fenix BoF

• verify VIM in production
• parallel testing
• OpenStack & Kubernetes master
• Raspberry PI

• python packages
• 10/10
• functional gates
• etc...

Kubernetes is becoming a site orchestrator choice to deploy network functions and application workloads in cloud-regions. ONAP is being enhanced to bring up VNFs and workloads as VMs and containers in K8S based cloud-regions. This presentation details following:

• Current K8S based use cases in industry
• NFV requirements
• Pre-work that was done in R3 to support K8S based cloud-regions.
• Architecture
• User Stories for Dublin Release
• vFirewall Demo

5GTANGO V&V integration with ONAP.

Highlight the existing test platform , V&V, in 5GTANGO that works currently with OSM and discuss potential for collaborative testing with ONAP and VNFSDK/ VTP in particular.

V&V will have the ability to invoke tests and publish results on ONAP Certified VNF Packaged

• Introduce current state of OVP and what comes in the future
• Demo of how to run some OVP tests (including OPNFV and ONAP)
• Show how to use the OVP web portal

Dublin Plans and beyond.

• Clover utilizes Istio, Prometheus, and Jaeger to provide container networking polices and visibility
• It also provides a cloud native network tracing tool to augment traces provided by Istio/Envoy
• Demo on visibility and network tracing to demonstrate how they can help in CI/CD pipeline

• Review vCPE use case implementation
• Discuss on how to separate use case custom code from ONAP platform

• Review Use Case
• Continue detail workflow and ONAP impact

Intel internal teams that use ONAP as users had good feedback and wanted to share same here

• Challenges faced during installation
• Challenges faced during usage
• Stability and performance challenges
• Suggestions for improvement

• Overview of the Edge Cloud Project
• StarlingX
• OVP Integration
• Resources

• License Hygiene
• Meeting/Calendar
• Better Alignment to open source norms

• Use a Sample VNF from OPNFV – L2 forwarder
• Show how we can do perf. benchmarking using OPNFV Yardstick NSB

• Onboard the VNF onto ONAP
• Deploy the VNF using ONAP onto an OPNFV OpenStack scenario

 Proposed VNF Badges for Dublin

Process to engage

• Overview of VF-C Casablanca Status
• Dublin plan

• Overview of Usecase-UI Casablanca Status
• Dublin plan

• Overview of CCVPN

• Usecase
• Demo
• Dublin plan

• Dublin extension of CCVPN Use Case (E-LAN services)
• Discussion on Implementation options

• Survey Report on SP Requirements for ETSI

• NFN Alignment
• Proposal discussion for Collaboration between ONAP and OPFV to help

• Requirement Analysis for Generic TOSCA Parser
• Gap Analysis for potential candidates and current implementation
• Proposal discussion for Collaboration between ONAP and OPFV to help

VNF Benchmarking with OPNFV Yardstick NSB(Network Services Benchmark) 

• Presentation/Intro
• Introductory lab where you will
  • learn how to install Yardstick-NSB on bare metal,
  • configure it and characterize NVFI and a sample VNF. 
  • Visualize results/KPI'

• • s

Alex Xia

Sriram Rupanagunta

• Use Affirmed Networks' MMC related VNFs and onboard them through SDC design tool

• Create a single Network service using these VNFs through SDC

• Perform SDN-C preload of VNFs

• Deploy Network service (MCC) using VID in ONAP portal

• Post-deployment configuration for the VNFs

• Define Edge Spec for Pharos lab
• StarlingX integration with XCI

• Experience of CMCC on edge cloud testing (OpenStack & Acceleration)

In 2015 and early 2016 the Docker community created the v2.2 image specification with a new type—a manifest list—to represent a list of platforms (OS, CPU architecture, etc) and references to platform specific layers for different runtimes.

In September 2017 Docker's Official Images became multi-platform. Running "run docker run hello-world" will pull and run the correct hello-world image whether you are using ARM, x86-64 Linux, Windows, IBM Z mainframes or any other system where Docker runs.

Because ONAP is vendor-agnostic, multi-platform support in the container ecosystem brings capabilities the ONAP community has been waiting for a long time.

The introduction of multi-architecture images will improve usability and deployability of ONAP containers across architectures and platforms. Developers no longer need workarounds such as adding an architecture-specific prefix to image names.

• Introduce the key concepts behind mult-architecture container images.
• Describe and discuss what it means for a container image (e.g. Docker) to be multi-architecture.
• Demonstrate how multi

• -arch images are built

SO Adapter which uses ETSI NFV SOL003 to connect VNFM

• Dublin Use Cases and Requirements
• SO VNFM Adapter Architecture Proposal

• overview of the status
• Dublin plan
• next steps for existing topics: 
  NS/ how to model slice profile.

There were official SDO coordinator officer link here: https://wiki.onap.org/display/DW/External+Standard+Coordination

We will give a yearly report and invite one SDO speech during the session which is most influential

Summary of ONAP Mini Summit organized in November in Warsaw.

We will share main outputs and takeaways

from the event.

As we know, HPA provides ONAP with capability of awareness of underlying hardware platform, and with that ONAP is able to leverage HPA features, such as DPDK and SR-IOV to locate and onboard VNFs on the right platforms for the purpose of performance mostly. However with maturity of ONAP and requirements of new VNFs, there exist some gaps inside the current HPA framework, for example, ambiguity tolerance in TOSCA templates and service assurance after VNF onboarding.

Apart from the above, taking edge computing or 5G as an example, we're facing more and more new workloads and new scenarios, where high bandwidth and low latency are highly required in the requirements. For instance, those VNFs which do real-time analysis and diagnosis on surveillance video stream need accelerators like GPU and FPGA evidently to achieve the goals, and should consider and plan corresponding life-cycle management, hardware metrics monitoring, QoS, etc. for the accelerators.

In the presentation, we're about to cover the following:

- Gaps between the existing HPA framework and the new requirements on accelerators

- Difference between HPA and acceleration management

- Framework proposal to support acceleration management in ONAP

- Changes to enhance ONAP and the HPA framework

- Interoperability with Multi-Cloud and VIMs

- Our plans

Presentation of the lesson learned from I phase of GNOCe  virtual robot implentation, ONAP PITEC, to monitor the quality of network management systems access in Poland, Slovakia, Belgium, Spain and Romania. This was possible thanks to the use of some of the capabilities of ONAP. In the first phase, ONAP is used to on-board and instantiate IPTEC probe VNFs, and to collect measurements results through VES collector and DMaaP.

In a  second phase, the use of ONAP should be extended to:

-       control loops development,

-       ONAP analytics integration,

-       PITEC re-configuration based on thresholds,

-       Integration with external ticketing system.

 NBI enhancement, service modification & notifications

Improving NBI TM Forum API for bss & oss means more integration with

• AAI
• SO
• SDC

Apis

Objectives:

• service modification ( get characteristics & values from AAI )
• more notifications ( from AAI, SO, SDC )

See

https://onap.readthedocs.io/en/casablanca/submodules/externalapi/nbi.git/docs/offeredapis/offeredapis.html

Present the evolution of platform component: Service Orchestration & Controller Design Studio; enabling self-serve to address pre and post instantiation configuration use cases.

OPNFV functional testing provides some trust indication on the major infrastructures (OpenStack and K8S).

As ONAP is now installed on top of K8S usually on top of OpenStack, it is important to trust the infrastructure hosting ONAP deployement before running deployements and tests.

OPNFV testing capabilities could be reused.

Additionnaly Pharos resources could eventually be used by ONAP on top of OPNFV scenarios.

Discuss the impacts of the Dublin release CCVPN Use Case on External API from both a Legato (Northbound: BSS/OSS to ONAP) and Interlude (East/West: ONAP to ONAP) perspective.

Also discuss MEF Interlude Alignment 

Could be done as a breakout.

Request Wednesday breakout session due to participant availability.

Update from SECCOM, including preliminary draft generic security requirements, milestone checklist additions, casablance in

review, Dublin requirements.

Janne Suominen

Mika Rautakumpu

Augmented Reality demo

on Edge Cloud

Augmented Reality demo

on Edge Cloud using

OpenEdge HW

5G Use Cases follow-up:

3GPP PM Data file collection

PNF sw upgrade

CM configuration via Netconf

1. Kubernetes = both VIM and NFM equivalent
2. KubeVirt to handle the "V" in VNF
3. An operator for "legacy" lifecycle management (OpenStack Heat compatible) = G-VNFM equivalent
4. Custom operators (e.g. Ansible operators) to allow for vendor enhancements = S-VNFM equivalent
5. Multus to attach the data plane (as well as the expanded control plane that some VNFs require)
6. A controller to manage these extra networks
7. ONAP on top for management and integration: OpenDaylight for SDN, TOSCA for stitching together network services, BPMN for OSS/BSS, A&AI, DCAE

The technological ecosystem is almost there, it just needs an additional nudge to bring the pieces together.

In this creative session, we'll identify security risks: threats and vulnerabilities for the assets in ONAP deployment. This is a continuing exercise from the F2F event in Beijing in June. The target of the assessment is a typical ONAP deployment with any needed assumptions on the deployment environment. Probably we need a few typical-ones like private cloud, public cloud.

After the risk identification, we'll do rating: risk impact & risk probability → risk level. The risks with unacceptable high level shall be mitigated, by adding items on projects' backlogs as necessary.

A fictious example, to give you a more concrete view:
-Asset: VNF O&M passwords stored in ONAP
-Threat (can be several per asset): unauthorized disclosure
-Vulnerability (can be several per threat): passwords included in plain text in a log file
-Risk mitigation: do not include any passwords in log files

For more info, please see Risk Assessment concepts, definitions, process: https://wiki.onap.org/download/attachments/11928162/

ONAP%20Attack%20Vector%20Analysis%20-%20Overview%20of%20the%20Process.pptx?api=v2

ONAP on Service Mesh

To ease operation of ONAP deployment, we explored running containerized ONAP components on Istio-K8S. By leveraging OPNFV Clover, we also gained more visibility into the deployment. In this presentation, we will present:

·         The general benefits of running microservice pattern application on a service mesh

·         How such deployment will benefit ONAP?

·         What is OPNFV Clover, and how it helps microservice based applications via augmenting Istio?

·         Demo

Regular Change Management meeting where we will discuss our plans for Dublin

1. E2E Workflow Designer/Orchestrator (Includes all existing Activities/BBs)
2. Schedule Optimization with Automated Conflict Avoidance
3. Traffic Migration Workflow
4. 5G PNF Software Upgrade

https://wiki.onap.org/display/DW/Change+Management+Dublin+Extensions

Full day hands-on session dedicated documentation. Open to work on or discuss,

• Project specific documentation
• Composite/overview documentation

Presentation about current state of ONAP offline deployments

https://wiki.onap.org/display/DW/Beijing+OOM+ONAP+Offline+Deployment

There are still manual steps involved in getting VNFs on-boarded  and during scale out operations.

This presentation and discussion is aimed to give a “consolidated” view of “all” that is required to achieve a zero touch or completely automated operations during service onboarding / instantiation, scale-out, scale-in operations and getting required configurations in place during this process.