Design And Implementation Of Basic Performance Evaluation Framework For IaaS Cloud Platform

With the rapid development of the cloud computing technology,more and more enterprises and users who benefit from the low cost,high quality,convenient deployment,and simple operation and maintenance of cloud services deploy services to the cloud.But in recent years,a series of frequently reported cloud service failures show that there are still some shortcomings in the current commercial cloud service technology.How to conveniently and scientifically evaluate the performance of different cloud platforms to meet the purpose of guiding ordinary tenants to choose cloud service,and to meet the needs of cloud providers to continuously optimize the performance of their own cloud platforms based on the evaluation results is urgently needed to be solved.In view of the above problems,a performance test framework for Iaa S cloud platform is designed and implemented in this thesis based on the research of existing cloud platform test theory and some mature test tools.The framework is simple to operate and the testing process is fully automated.The framework also can test different performance evaluation indexes of different cloud platforms,as well as the performance evaluation indexes are easily extended by developers.Aimed at testing the performance of cloud platform,the test scheduling management,load generation,cloud platform cluster monitoring,data collection,data analysis,report generation and other technologies in the test all are researched.The framework also provides a tool set and SDK for users to develop automated test scripts so that developers don't need to care about the internal details of the technology.Firstly,this thesis introduces the research background,the research significance and the research status at home and abroad.Then,the key technologies used in the framework,such as cloud computing,cloud platform performance test Open Stack which is a open source Iaa S cloud platform,cloud platform cluster monitoring tools,load concept and the technology of report generation are briefly presented.Next,according to the purpose of cloud platform performance test,the functional requirements of the system are sorted out,the overall architecture and the network topology of the cloud test framework are proposed,the design of the performance automation test flow and the design of the module interaction process are introduced emphatically.Based on the above design,the framework is implemented on the Open Stack.The implementation consists of many features,e.g the general workload,load driver,cloud platform cluster monitoring,data collection,data analysis,test driver module,the content of report in the cloud platform test,report generation technology and user interaction module.Along with the achieving process,cloud platform dynamic scaling test is used as an example to describe the implementation details.Benefiting from Python scripts' cross-platform and migration easily,cloud platform performance automation test is mainly done by some test scripts.Finally,the deployment of the test framework and the tested cloud platform are introduced.The use of the test framework is briefly explained.Several test cases are designed to verify the overall features of the test framework.Through the test cases,the cloud platform basic performance test framework has been proved to be effective and scientific.