A microservice fault identification method based on LightGBM

With the development of cloud computing, the software system architecture has gradually changed from a single architecture to a service-oriented architecture, of which microservice architecture is a typical representative. It is committed to providing users with more reliable, maintainable, and extensible software design services. Although the microservice architecture has many advantages, because there are multiple services in the microservice architecture, it becomes more difficult to detect faults when the system fails. How to efficiently detect the causes of faults is the key technology to ensure the performance and reliability of microservices. Aiming at this problem, this paper proposes a microservice fault identification method based on LightGBM method, which can analyze the historical operation information of microservices, learn and locate the fault causes, be used for fault identification, can quickly locate faults, and ensure the high availability of microservices. Compared with GBDT and XGBoost methods, the experimental results show that the accuracy of this method is 0.85, the recall rate is 0.81, and the F1 score is 0.83. Compared with other fault detection models, this method improves and can effectively detect abnormal services and identify fault microservices.