Microservice Patterns for the Life Cycle of Industrial Edge Software

Abstract

The Industry 4.0 initiative is pushing traditional production systems and methodologies through fundamental changes. In traditional industries such as manufacturing, utilities, and infrastructure monitoring, customers are demanding more flexibility and greater efficiency. The software that oversees these systems addresses these demands by adopting novel architectural styles and patterns. Industrial edge systems are gaining momentum because they offer superior data security and better response time to field operations. Microservice style has been adopted from cloud services because it improves development efficiency. This paper presents architecture patterns that cover four phases in the life cycle of industrial software. In the deployment phase, DOWNSHIFTING moves microservices from the cloud to the edge. During runtime, MONITORING CONNECTORS oversee properties of microservices at the edge, such as performance and resource usage. When adaptation is needed, EDGE-TO-CLOUD LOAD BALANCING leverages cloud resources to fulfill performance expectations. When a new microservice version is deployed, TEST PROBE enables testing the microservice at the edge without interfering with ongoing operations. This paper is intended for architects, developers and practitioners who are involved in delivering software in industrial edge environments.