Research on Microservice Application Performance Monitoring Framework and Elastic Scaling Mode

Abstract

Abstract Compared with the traditional monolithic architecture, the microservice architecture style divides a system into different microservices which run in the distributed system. The complex dependencies between services bring new challenges to the monitoring analysis and quality assurance of system performance. According to the characteristics of microservice application, a performance monitoring framework based on big data is designed in this paper. It monitors and controls the microservice performance through data collection, big data storage, elastic scaling management, integrated scheduling and so on. Furthermore, an elastic scaling mode based on time sliding window and scene driven is proposed. Experiments show that this mode could realize resource expansion prediction and resources saving. This research is helpful to real-time monitoring and continuous optimization for microservices, which will effectively promote the integration process of development, testing and maintenance for microservice application in SGCC (State Grid Corporation of China).