The impending launch of 5G telecom networks is generating excitement, promising to fuel innovative technologies and introduce new use cases and architectures that will enhance our digitally connected world. A pivotal feature of 5G is edge computing, which will revolutionize how devices interact with centrally located data centers. Ultra-low latency, higher bandwidth, and secure data transmission are core requirements for upcoming use cases such as augmented reality (AR), virtual reality (VR), connected/autonomous cars, and the Internet of Things (IoT). The network edge will address these demands. So, now let us see How Can Serverless at the Edge Resolve Resource Utilization Concerns along with User-friendly Mobile Network Monitoring Tools, Mobile Network Drive Test Tools, Mobile Network Testing Tools and User-friendly LTE RF drive test tools in telecom & Cellular RF drive test equipment in detail.
Understanding the Network Edge
The network edge is envisioned as a mini data center physically close to digital devices, where data generated by devices like sensors, smartphones, autonomous cars, and industrial equipment is processed, aggregated, summarized, and analyzed. This refined data is then forwarded to the cloud, significantly reducing the volume of data sent to central data centers. Consequently, this minimizes network traffic, reduces response time, and lowers transmission costs.
The Problem
Despite its potential, edge computing faces significant challenges in delivering services with dynamic performance and scalability. Edge nodes vary in size and resource requirements. Some nodes are small and dedicated to specific tasks, while others may be used for telecommunications or specific events like sports stadiums. These edge nodes are spread across thousands of locations and connected to central clouds or data centers. The workloads on these nodes will grow as the number of connected devices increases. Additionally, edge infrastructure comprises diverse software and hardware components from various vendors, necessitating continuous integration, monitoring, and updating to support new services for end consumers. The edge thus becomes a highly networked set of mini data centers, each physically close to digital devices and customers.
AT&T’s Edgility: Bringing Serverless Capability to the Edge
To address these challenges, AT&T developed an initiative called “Edgility.” This initiative aims to implement serverless functionality at the network edge, efficiently utilizing computing resources to handle event-triggered use cases like IoT.
Architecture of Edgility
A 5G network based on edge architecture includes a central cloud/data center and hundreds of connected edge nodes. In the Edgility architecture, each node is based on the Akraino Edge Stack, an AT&T-led Linux Foundation project. The Akraino Edge Stack offers a cloud-native approach, based on microservices architecture, enabling workloads to be managed using both containers and virtual machines for reliability and high performance.
The major component in this architecture is Kubeless, a Kubernetes-based serverless framework integrated with the Akraino Edge Stack to enable function as a service (FaaS). This setup allows serverless functions to be deployed on top of Kubernetes clusters, leveraging the advantages offered by Kubernetes.
Functionality and Benefits
The Akraino Edge Stack provides a microservices architecture in which workloads are deployed mainly in containers with smaller computing footprints. Integrating FaaS through Kubeless further advances this architecture by breaking microservices into smaller functions, which can be called upon when specific events are triggered. Until then, these functions remain idle, optimizing resource utilization at the edge and ensuring active services have sufficient computing power.
Contributions and Integration
AT&T’s contributions to the Akraino Edge Stack include integrating components like Kafka, Message Queue IoT Gateway (IOTG), Kubeless, and Data Plane Kubernetes. These integrations enhance the serverless edge node capabilities, allowing for efficient processing and communication of IoT events and other edge workloads.
Orchestration with ONAP
AT&T’s open-source contributions to the Linux Foundation include the Open Network Automation Platform (ONAP), which is used for the design, deployment, and orchestration of the Akraino Edge Stack. ONAP’s Service Design and Creation (SDC) tool creates blueprint models, while Cloudify and TOSCA serve as service orchestrators for deploying these models across multiple edge stacks. Clustering of functions at the ONAP level and the edge node level is managed using Kubernetes, with monitoring and visualization handled by Grafana and Prometheus.
Predictive Operations in the Second Phase
The first phase of Edgility successfully tested deployment and operations for real-life use cases like Intelligent Transport Systems. The second phase will integrate artificial intelligence (AI) and machine learning (ML) capabilities through the Acumos framework, another Linux Foundation open-source project. This integration will enable predictive orchestration of service functions and dynamic resource utilization based on ML analysis.
Business Advantages of Edgility
The extensive and complex 5G network, with its thousands of edge nodes, poses challenges in terms of capital expenditure (CAPEX) for service providers. Maintaining high-performance infrastructure at the edge increases costs, while limited IT resources and computing power raise concerns about service availability and reliability. Edgility addresses these issues by adopting the serverless or FaaS methodology for edge application services.
Serverless technology, already embraced by enterprises for cost savings in public clouds, charges only when a function is called and resources are used. Similarly, Edgility optimizes resource utilization by activating services/functions only when needed, keeping inactive services idle. This approach creates an event-driven, high-performance platform where microservices and their functions can utilize both public clouds and external data centers. Service providers gain control over function workload hosting, reducing CAPEX investments and benefiting from cloud-native features.
Summary
Edge computing is essential for enabling most 5G use cases. However, the resource requirements and expectations from edge nodes present significant challenges, especially with the growing number of connected devices and data. Edgility addresses these concerns by offering a serverless edge node solution, ensuring computing resources are utilized and optimized to meet varying data processing demands. This approach not only improves performance but also reduces costs, paving the way for the efficient deployment of edge computing in the 5G era. Also read similar articles from here.
