Microservices Containerization in SBCs (Single Board Computers): A Cloud Edge Computing Approach

With the recent and unprecedented increase in demand for Cloud services, furtherly promoted by 5G, Edge computing is emerging as an indispensable technology. Tailored to mitigate the continuously growing load on Cloud data centers and cope with the rising proliferation of IoT (Internet of Things), Single Board Computers (SBCs), embedded systems, and microservices-based applications, edge computing is turning into an integral technology enabler in 5G. Arguably, most of edge microservices will be deployed using virtualization, and specifically using containers instead of VMs (Virtual Machines). Dubbed 5G-MEC (Multi-Access Edge Computing), the 5G edge has to cope with 3 major services: eMBB (enhanced Mobile Broadband), mMTC (Massive Machine Type Communication), and URLLC (Ultra Reliable Low Latency Communication). In this paper, we shed further light on the fundamentals of cloud edge computing and present the subtleties of deploying a real-world SBC-based distributed edge application. The latter is an AI-based application, embedding an image recognition microservice running in containers, deployed in Raspberry PI SBCs, and orchestrated using Kubernetes.