Design And Implements To Job Scheduling Of Code Defect Test System

In the process of the research and development of software,code defect testing provides an important guarantee for the quality and reliability of the software.With the increasing popularity of cloud computing technology,migrating the services of code defect testing to the cloud platform can ully utilize computing resources of the server cluster.It will meet target users for the code defect detection requirements efficiently and quickly.DTS code defect test cloud platform comes into being based on the above concept.In current version,it is deployed relied on virtualization technology and Hadoop distributed system infrastructure.Each server runs code defect testing separately.Because each server only test a complete project,it's just the succession of standalone mode in terms of running mode,and fails to make full use of cloud computing technology to add the strong computing power of the server cluster.Due to differences of numbers and sizes of the project,this running mode causes part of the cluster server idle,and results in unbalanced load of server cluster.Aiming at this problem,this paper presents a new job scheduling strategy on DTS code defect test cloud platform.This strategy unpacks the test project uploaded by user into source files as the basic test unit,and every source file is tested as a task.The algorithm of job scheduling on the job scheduler chooses test task.The chosen task is dispensed to a server and finishes testing.Hadoop MapReduce computing model collects test results of a project and ultimately generates complete test results.According to the organization of the web application.DTS code defect test cloud platform is implemented based on Struts2 framework and Spring framework.It uses MVC design pattern.Through comparing and analyzing classical job scheduling algorithms and distributed job scheduling algorithm,we conclude a task distribution strategy of a job scheduler which is suitable for code defect testing service.The author verifies the application of this strategy on DTS code defect test cloud platform through related experiments.The experimental results show that the job scheduling strategy proposed in this paper has a significant effect on the performance of the code defect detection service provided by DTS code defect test cloud platform.