Website performance study

• Main Author: Qian, Lei
• Other Authors: Althea Liang Qianhui
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Computer science and engineering::Computer systems organization::Performance of systems
• Online Access: http://hdl.handle.net/10356/70547
• Institution: Nanyang Technological University

This project presents an analysis of multiple techniques that are used to improve the performance of web applications. Taking into consideration customers’ high expectation of web application performance and potential negative consequences brought by poor performance, this report offers a comprehensive range of solutions to enhance web performance so as to perfect end-user experiences and further boost business bottom lines where applicable. The experiment was carried out on the self-built blog web application. The HTTP requests used for the experiment were chosen based on their degree of representativeness of the website. Dedicated techniques were applied to each request. The impact of those techniques on the performance of the web application was measured by Apache ab test. In order to prevent network bottlenecks, Apache ab test was carried out within the intranet of the server. Database based techniques were proven to be one of the most effective strategies when it comes to improving website performance. Based on the usage of the web application, an effective caching design was able to prevent repetitive database access so as to improve the performance of the web application to a very large extent. By scaling out and deploying a load balancer, the performance of the web application increased linearly with the number of servers. Moreover, client-side technique AJAX was able to minimize the amount of data transfer between the clients and servers, allowing servers to serve more connections simultaneously. Understanding these solutions is the key to implementing responsive web applications. It recommends that further studies and experiments could be carried out by implementing these techniques on distributed systems.