Research On Web Service Performance Assurance Technology In Cloud Environment

With the in-depth development of cloud computing technology and its application, Web Service carries background processing tasks of many network applications. The mode based on Web Service has become the preferred method of developing Internet application system. In the way of the Web Service development, whether it is based on traditional SOAP-architecture Web Services or today's REST-architecture Web API, their essence is a kind of network data interface. Web service performance influences directly the performance of the application.This theis starts from the requirement of Web Service performance guarantee, sums up the key problems in two aspects: to reduce the Web Service's "request response delay", and to improve the running efficiency of Web Service.In view of the above questions, this article focuses on "workload scheduling" and "computing performance", the two core aspects of cloud computing, and proposes a performance guarantee model for web services in cloud computing environments using the idea of SDN(Software Defined Network) and the technology of virtualization. Our work is to solve the problem of Web Service's “failure” caused by poor performance at the peak time. The main work in this paper are as follows:1. Service performance guarantee model in cloud computing environmentThe performance guarantee model based on cloud environment is divided into two parts: performance monitoring and performance control. Performance monitoring focuses on real time monitoring of web service, provides data basis for performance control; performance control focuses on scheduling Web Service request data and adjusting the resources that the web service needs.2. Research on Web Service load scheduling technology in cloud computingenvironmentThe research work of Web service load scheduling technology includes two aspects.(1) The first one is the priority scheduling technology to address the different types of service scheduling problem in case of request congestion. This technology uses Click software router to support dynamic allocation and schedule service priority. With the software control module to adjust the performance of priority, it can better protect critical services.(2) The second one is the load balancing technology to schedule requests with high concurrency. This technology is based on the traditional LVS(Linux Virtual Server) model of load balance. It combines with the software defined network in the cloud environment and transfers the response data return from the other way in the purpose to make the scheduler mainly focused on the real server selection decisions to increase the load efficiency.3. Research on the security of Web Service computing resources under the cloud computing environmentThis part begins from the underlying calculation of Web service program, and studies the computing performance and resource management of service program.(1) The work on service computing performance is mainly about external data I/O performance optimization, i.e., to improve the response speed of service by improving the service implementation process on the external data I/O performance.(2) The work on service computing resources is mainly about dynamic prediction and control of physical resources required for computing services(CPU, memory, etc.), using performance monitoring data and elastic control technology based on cloud, to better guarantee the physical resources for web services.4.Implementation and verificationThanks to Aliyun platform, we can handily create an application on the public Internet based on Web Service architecture. Through this application, on the one hand, we use the technology mentioned in the thesis into the performance security of Web Service. On the other hand, with Aliyun's open interface called “Fei Tian” platform, the resources of Web Service can be easily controlled. At last, through the real running environment, the thesis' s technology is verified to be effective in the cloud computing environment.