Research And Implementation Of Scheduling Optimization System Based On Application Behavior Pattern

Application performance management is becoming more and more important for monitoring system performance and improving system robustness.However,monitoring system can only reduce the negative impact.It is more important to analyze the system and make reasonable suggestions for improving system performance.With the complexity of large-scale applications and the development of hardware technology,server resources become more and more abundant.Scheduling projects when sharing server is of great importance for improving system efficiency and improving server resource utilization.To solve the above problems,this paper design a scheduling optimization system based on application behavior pattern.First,mine application behavior patterns by applying historical data,and then by analyzing behavior patterns,selecting strategies to schedule applications.In this way,the efficiency of the application and the utilization of the server resources are improved,by optimizing the scheduling of multiple applications in the server.This paper first introduces the related background and technology.Then,according to the needs analysis of scheduling optimization system,we design an application behavior pattern mining algorithm based on clustering algorithm and a scheduling algorithm based on application behavior pattern.Then design the overall structure of scheduling optimization system,combined with the overall framework of APM application performance management system and scheduling system work scene.According to the functional requirements,the system is divided into modules,and the specific demand analysis and function decomposition are carried out for each function module.Meanwhile,the interaction design of each module is given.And then,this paper introduces the internal and external interface and data structure of the scheduling optimization system.Finally,the system is designed in detail,and the key process to realize the system is given.And the development and realization of the system are completed.In the end,the function test of the system is carried out,and the test cases and testing processes of core function modules are introduced in detail,which verifies that the scheduling optimization system basically realizes all the functions needed.