Performance Optimization Of Web Applications

With the rapid development of Internet, Web-based applications are more closely together with the way people live and work. Many traditional information systems were ported to the Internet. Microblogging, e-commerces, social networking sites; these new applications are more and more. Because the Web is interactive, convenience, ease of use, etc., and the many businesses and individuals are like it. Performance is a decisive factor in the success of the Web application. Therefore, to optimize the performance of Web applications has become an important research issue.Usually, when we talk about a performance of Web applications, when thought is to optimize server performance backstage, front-end performance optimization is often overlooked. Currently, the back-end Web application performance optimization technologies has been very mature, such as load balancing, reverse proxy cache, distributed cache, shared file systems, performance optimization of databases, and so on. But when the rise of Web 2.0 technology-based Web applications, a common phenomenon is that, in real applications, the end user perception of this application is very slow response, and in the system server side, but can not find any resource bottlenecks. With the arrival of Web2.0 era, the contents of a page containing more and more, more and more abundant. Get a page, HTTP requests average over 90, the average size of the page is more than 1.5M, the average need to request to 15 domains [1]. Therefore, the front-end Web application performance optimization can no longer be ignored.Firstly, we study the computational model browser response time, but also analyzes the HTTP protocol. Meanwhile, the analysis presented in user-perceived latency and performance as measured by indicators, in order to verify the performance of the program as a follow-up measure. Then, based on the number of HTTP requests Factors, rendering speed, caching mechanism was proposed to reduce HTTP requests and speed up the rendering speed, set the caching optimization program, and gives the specific implementation and verification, to confirm these optimization strategies can indeed reach performance optimization purposes.Furthermore, this paper focuses on the problems of processing Web requests. Traditional Web requests handled cause the presence of the user experience and performance poor. The AJAX technology is widely used, although the improved performance and user experience, a large number of requests has increased, resulting in a heavy burden on the server. Therefore, this paper proposes a Web request pipelining solution.Request pipelining is to take advantage of the parallelism between the Web server and browser, first decomposed into multiple pages can be called small Pagelets, then build the pipeline through a Web browser and management server and run them at different stages. Generally, the use of multi-threaded pipelined. Here, the proposed single-threaded implementation of request pipeline and PageLet priority. Through experimental analysis, confirmed the multi-threaded pipelined manner do better than AJAX performance in terms of performance, and the relatively light burden on the server. However, in the face of a large number of concurrent requests, it will also cause a serious decline in performance; while single-threaded implementation, relatively small performance impact, and single-threaded implementations can also get a good user experience.Therefore, proposed an improved Web requests pipelined. That is, by adding a layer scheduling server, under normal circumstances, to schedule multi-thread module; in the case of a server overload, to schedule single-threaded module. Meanwhile, through experimental analysis, a good performance improvement can be obtained.