Microservice Container Load Trend Prediction based on Informer and Batch Normalization

The microservice container load refers to the workload and resource consumption carried by the microservice during operation. With the increasing scale of microservice application systems, the existing passive cloud resource management and control model has been unable to meet the needs of large-scale microservice operation and MAINTENANCE. To address this issue, we introduce a load trend forecasting model tailored for microservice containers, which leverages the Informer model in conjunction with Batch Normalization. Our proposed approach demonstrates significant efficacy in precisely forecasting container loads. It adeptly resolves the persistent issues associated with missing data and the efficiency limitations encountered when dealing with large-scale data, which have posed challenges for existing models.