A Software-defined Edge Computing Solution

Project Information

ONEedge

Grant agreement ID: 880412

DOI

10.3030/880412

Project closed

EC signature date 30 September 2019

Start date 1 November 2019

End date 31 March 2022

Funded under

• INDUSTRIAL LEADERSHIP - Innovation In SMEs

Total cost

€ 2 143 735,00

EU contribution

€ 1 500 614,50

Coordinated by

OPENNEBULA SYSTEMS SL
Spain

Periodic Reporting for period 2 - ONEedge (A Software-defined Edge Computing Solution)

Reporting period: 2020-11-01 to 2022-03-31

Summary of the context and overall objectives of the project

An increasing number of organizations and private companies—from a broad range of sectors like gaming, smart cities, the Internet of Things (IoT), social networking, government, and telecommunications—are currently focusing their technological and business strategies on being able to provide innovative services and capabilities with absolute immediacy to their customers and end-users. In these cases, latency is the key factor for quality of experience. “Edge Computing” is an increasingly mature paradigm based on the concept of shifting away resources from centralized cloud infrastructures towards carrying out the processing as close as possible to mobile devices, sensors, and end-users. Yet, despite the urgent need to define the foundational platform to deliver edge services, and the dramatic growth in industry investment, companies that are currently developing these next-generation solutions just do not have a viable and cost-effective platform on which their low latency applications, running on virtual machines or containers, can become mainstream.

The ONEedge project has expanded the OpenNebula software with new innovative edge computing features that provide companies with an automated software-defined platform to easily build their own private edge cloud environments based on the aggregation of highly-dispersed edge nodes from existing infrastructure providers (including hyperscalers and telecom operators) in close proximity to end-users, devices, and data sources. ONEedge enhancements have been also delivered as part of Edgify, a new standalone Edge-as-a-service distribution that provides an On-demand Edge Cloud Service.

With Edge Computing having been identified as one of the major disruptors in the cloud and data infrastructure sectors, it became crucially important to make sure that new technological innovations, especially those that provide a significant competitive advantage, were really accessible to as many companies as possible. ONEedge, as an open source project that will lower the barriers of adoption for both large organizations and SMBs across and beyond Europe, encapsulates that principle and provides a safeguard against edge computing becoming a technology only accessible to large corporations.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The work carried out during the ONEedge project has been focused, not only on developing a first version of a sophisticated edge computing platform to innovatively address the needs for lower latency and provide a simple, distributed infrastructure management, but also to generate actual demonstrations of its capabilities in real life situations, as well as to get some of these capabilities in the hands of users to put them to the test by themselves. As this is an open source project, a great deal of our work managing the project is not only about producing high-quality code, timely deliverables, and informational materials, but also to coordinate the creation of these items and to make them really accessible to the broader public.

- Definition and incremental updates of the ONEedge Architecture with the main components of this innovative edge computing platform, and the main software requirements derived from user requirements.

- Release of four versions of the innovative architectural components as part of the standard OpenNebula distribution: OpenNebula, 5.12 ‘Firework’, OpenNebula 6.0 “Mutara”, OpenNebula 6.2 “Red Square”, and OpenNebula 6.4 “Archeon”.

- Release of Edgify, a new standalone edge-as-a-service distribution that provides an On-demand Edge Cloud Service where companies can easily build their distributed edge environments using resources from existing cloud/edge infrastructure providers.

- Development of use cases in relevant sectors, such as Gaming, IoT, Streaming, Telecommunications, Smart Tourism, Smart Wildfire Management, Smart Cities, and Farming.

- Continuous monitoring and evaluation of the edge market dynamics, community building, user engagement, and business plan definition.

All these results are described in the project reports that are publicly available for download from OneEdge.io

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The work carried out during the project has had a special focus on leveraging OpenNebula by expanding this well-established European open source cloud technology with new edge computing features and integrations, and building and delivering the new standalone Edge-as-a-service distribution Edgify. Main results include:

- New innovative highly distributed Edge Cloud Architecture consisting of an Edge Cloud Management Cluster and an Edge Cloud Infrastructure made of Edge Clusters that can be automatically deployed both on premise and on public cloud or edge providers to enable true hybrid environments.

- Automatic Edge Cloud Management Cluster deployment, configuration and bootstrapping, and complete life-cycle management of the ONEedge front-end instances.

- Provisioning tools and first version of Edge Provider Catalog to automatically deploy and scale Edge Clusters in cloud and edge data centers with provider drivers for main cloud and edge providers, and on-prem central and edge locations.

- Edge Cluster deployments with hyperconverged (HCI) architecture, datastores based on a new 3-tier replica cache storage datastore or Ceph storage, new drivers for IP address management, private networking based on VXLAN and EVPN BGP inter-edge extensions, and innovative backup interface.

- Automatic deployment of Kubernetes clusters with scale-in and scale-out capabilities, execution of clusters across nodes in a highly-distributed edge environment, persistent volumes and dedicated storage nodes, and multi-master configuration with an associated load-balancer.

- Integration with Firecracker and container marketplaces, a new open source virtualization technology based on microVMs especially designed for creating and managing secure, multi-tenant container and function-based services.

- New application workflow orchestration with deployment dependencies and auto-scaling rules with improved scalability and response time that allows complex multi-tier applications composed of multiple containers to be easily defined, deployed, and scaled up/down.

- Development of fine-grain NUMA placement to optimize the performance of workloads, especially VNFs, by allocating dedicated memory blocks to an instance, and new drivers to define virtual networks and security groups.

- New secure and scalable distributed control and monitoring system, able to scale to thousands of physical hosts and tens of thousands of VMs, with better support to hypervisors at cloud/edge locations and HA deployments.

- New Graphical User Interface, Provision FireEdge, for Edge resource Provision, and new Sunstone GUI prototype built using React/Redux and delivered by the FireEdge server.

The ONEedge platform provides a completely unique model to build Edge-as-a-Service environments that connects corporate users with existing and new cloud/edge infrastructure providers, including local data centers, hyperscalers, and 5G/telecom operators. Our solution creates a neutral two-sided marketplace that fosters new customer-provider relationships, minimizing vendor lock-in, bringing more balance to the cloud market, reducing entry barriers, and facilitating the emergence of new business models.

Distributed Edge Computing