Collaborative Deployment and Routing of Industrial Microservices in Smart Factories

In large smart factories, massive microservices compose complicated modular IT systems, providing various service functions. However, large microservices-based IT systems incur sophisticated communications and invocations among the massive microservices, which calls for efficient orchestration techniques to meet the high requirements in smart factories. Also, complex data interdependencies among microservices tightly couple deployment with routing, further intensifying the difficulties in orchestration. Such challenges demand delicate joint optimization of service deployment and request routing, which however, are neglected by previous work. In this case, this article investigates the collaborative optimization of microservice deployment and routing in smart factories. First, we construct a communication queuing network model to analyze service performance under dynamic load. Second, two heuristics are proposed to provide differentiated deployment and routing schemes for various demands. Finally, rigorous experiments validate that our approach significantly enhances network efficiency across various production scenarios in smart factories.