LF Networking Demos
Demonstrating Service Provider Specific Test & Verification Requirements (OPNFV, ONAP) - Proposed By Vodafone and Demonstrated by Vodafone/Huawei/iConectiv
This demo aims to demonstrate the capability of ONAP On-boarding process (via SDC) to interact with VTP (VNF Testing Platform) and request performing specific test scenario on VNFs before on-boarding them. The demo will use Standard, and OPEN API to interact with VTP which will also make it possible, in future, to interact with Certification Repositories (such as OVP) to retrieve a list of Certification a given VNF has undergo. This integration will open the door for more E2E Compliance and Certification process under a dynamic DevOps environment. VTP as a platform allows the interaction with OPNFV testing tools (such as Dovetail) and also with third party testing tools (or third party labs) that can carry operators specific test scenarios.
Network Service Mesh is gaining a lot of traction in the CNCF world and is disrupting the cloud-native networking. This demo intends to showcase integration of ODL and NSM through gRPC mechanism. It will showcase how NSM can leverage ODL network models and also how the integration can address many of the hybrid network scenarios.
SDN Performance Comparison Between Multi Data-paths (Tungsten Fabric, FD.io, Red 5) - Presented by ATS, and Sofioni Networks
This demo will show observations on the performance of Vector Packet Processing (VPP) versus Data Plane Development Kit (DPDK) vRouter, both controlled by the same Tungsten Fabric controller. It will have a Tungsten Fabric Control Plane connected to one of the compute running VPP as data path instead of the regular kernel module or DPDK vRouter, with all servers connected to the same Top of the Rack to ensure similar environment. It will demonstrate the side by side performance of the two data planes by adjusting the network conditions through background flow generator by changing the number of background flows injected into the network.
The BBS use case team will demonstrate its first implementation for ONAP Dublin release using ONAP for the design, provisioning, life-cycle management and assurance of broadband services. The use case covers new scenarios, such as nomadic ONT (Optical Network Terminal) and service subscription plan changes, and shows the extensibility of the ONAP platform in supporting the orchestration of services across different locations (e.g., Central Office, Core) and technology domains (e.g., Access, Edge) within the locations. Note: BBS implements/tests some of the specifications defined in the architectural framework of CloudCO (Cloud Central Office), Technical Report TR-384, among others.
OpenStack Rolling Upgrade with Zero Downtime to Application (OPNFV, OpenStack) - Presented By Nokia and Deutsche Telekom
In this demo we will make rolling OpenStack upgrade by utilizing OpenStack Fenix workflow through OPNFV Doctor test framework. With VNFM interaction, we can also guarantee a zero downtime to VNF. The Demo will aim to have the Fenix rolling upgrade CI job implemented, so we can also witness how modified DevStack scripts are used in upgrading the nodes. As result, we will be giving great tools for anybody to handle and test the OpenStack upgrade.
Skydive & VPP Integration: A topology analyzer (FD.io, Ligato, Skydive, Docker) - Presented by PANTHEON.tech
Skydive is an open source real-time network topology and protocols analyzer that aims to provide a comprehensive method for understanding what is happening in the network infrastructure. This demonstration contains running Vector Packet Processing (VPP) instances, using Ligato VPPAgent Docker containers, to automatically start the VPP instances and detect changes to the network topology in real time. The importance of this demo is based on the need to extract data from VPP and provide a birds-eye view of the network via Skydive.
VSPERF: Beyond Performance Metrics,Towards Causation Analysis (OPNFV) - Presented by Spirent Communications
This Demo will show how VSPERF and other tools to collect, analyse and visualize various result-data for each and every testcase, to facilitate causation analysis. The advantage of VSPERF architecture is having uniform timestamps across all result-data while measuring different configurations that supports correlation and causation analysis. For every test run, we plan to demo the following:
1. Heat-Maps of Correlation
2. Trend Analysis of Live-results
3. Trend Analysis of the EOT Results
4. Depiction of Bottlenecks in Topology
5. Detailed statistics of Critical Metric
6. Test-Environment Representation: Available, Provided, Configuration, etc.
7. Representation of all the events during the test-run
This ONAP demo will leverage many components/functional elements of ONAP architecture to deliver an e2e network slice with specific characteristics in an MBB (Mobile Broadband) services use case. It will instantiate and configure several virtual network functions and demonstrate the entire life cycle of service creation, instantiation and configuration of service. The demo will configure remotely an AT&T/Ericsson network in New Jersey using ONAP components and show the LCM of 5G network and create e2e Network slice.
Service Mesh and SDN: At-Scale Cluster Load Balancing (Tungsten Fabric, Istio, Consul, Envoy, Kubernetes) - Presented by CloudOps, Juniper and the Tungsten Fabric Community
From service mesh, to ingress, to network policy, to encryption and more, networking and security has shifted left, propelled earlier into the DevOps time horizon. In this demo, you’ll see how Tungsten Fabric, an SDN for Kubernetes, serving well beyond the basics of CNI, makes a great foundation and complement for service meshes like Istio and Consul. This demo will showcase Tungsten Fabric’s cloud-agnostic service load balancer implementation for Kubernetes, as well as compare and contrast Tungsten Fabric’s integration with Istio and Consul in the context of handling both legacy and cloud native workloads. We’ll also discuss scale-out load balancing, from gateways, to ingress, to service proxies.
The LaaS 2.0 project is intended to support the LFN community and is currently being used by the OPNFV and ONAP teams. With the launch of LaaS 2.0, open source networking project leaders can now book multi-node resources, proving their team with resources to deploy larger projects across multiple bare-metal servers. The booking process has been updated to allow the configuration of different OS images and networking between the servers, creating a flexible tool that can be leveraged by the LF Networking projects. This demo will help to let the communities know about the available resources and show users the booking process and the backend tooling. See the the LaaS dashboard here: https://labs.lfnetworking.org.
The Cloud Native Computing Foundation (CNCF) and LF Networking have partnered to demonstrate a path for the industry to move from Virtual Network Functions (VNFs) to Cloud native Network Functions (CNFs). Come visit us at booth #419 to learn about how you can get involved in the movement with the CNF Testbed.