Workload Modeling for Microservice-Based Edge Computing in Power Internet of Things

Microservice-based edge computing is a key technology to support power system to adapt to the high concurrency and diversification of application in power Internet of Things. Workload modeling is not only the basis for analyzing the logical design and computing resource requirements of application, but also the premise of optimizing the computing resource configuration and scheduling of edge computing. In this paper, a workload modeling method for microservice-based edge computing is proposed. The composition of an application and the logical relationship among the microservices are elaborated. Then, the topological structure graph and model are proposed to describe the logical relationship of the microservices. Based on the established model, the workload model is formulated to calculate the computing resource demand and latency of the application. Besides, a microservice-based workload evaluation algorithm is proposed to obtain the configuration results and the latency performance. The influence of the application architecture, topological structure of microservices and virtualization technology of edge computing terminal are analyzed. Simulation results verify the validity of the proposed model and method.