The Mobile Application Research About Services Performance In Cloud Computing Platform

This article discusses the background and the current research status about the selected subject firstly. It introduces the key technology used in the process of system development secondly. Then it analyses the main functional requirements about the system and the analysis process about Web application`s performance. Then the system is carried on the summary design and detailed design from the perspective of software implementation. Finally the system can analyze and manage the Web application`s performance based on mobile application under cloud computing platform. This system can analyze the web application which is added in the cloud computing environment according to the threshold of performance which is set by user. The results of this analysis are given when the analysis task is completed. Users can obtain all of the web application after they log in this system. Then users can analyze the selected web application timely. So they can understand the performance characteristics and bottlenecks of the selected web application based on the results of analysis. This also provides the reference of how to optimize the web application performance further. This system can inform users about the changes in the performance of web applications through the message push mechanism based on Android platform when web applications` performance change. This system is designed to provide users with good interface and information about web applications` performance timely. At the same time it improves the working efficiency of the cloud computing platform users.This system is developed based on C/S structure. Server implements the fine-grained monitoring of web application services through Nagios monitoring mechanism. Monitoring data is stored using MySQL database. On the basis of the performance monitoring, we implement a set of performance analysis method according to the characteristics of the web application through Bayesian classification technology which belong to machine learning. The main idea of analysis is:Firstly we should put forward a expectations based on the overall performance of web applications. Due to system performance metrics which have been monitored are associated with the overall performance of web applications. The final system performance metrics which have the greatest impact on the overall performance of web applications will be found through feature selection. In the process of feature selection we evaluate the characteristics of selected attributes by the accuracy of Bayesian classifier which is generated by attributes subset. Mobile client based on Android platform mainly adopt UI development technology. The main display interface uses Fragment technology. The network interaction and parsing is executed in the Service. And we manage the structure of code using MVP design patterns. Network data interaction is guided by the HTTP protocol. Data push function is implemented through network application framework which is based on Apache Mina.