Research And Implementation Of Front-End Performance Optimization Based On SaaS Platform

With the rise of SaaS model at home and abroad, more and more SMEs have experienced the convenience which SaaS model provides. Due to the continuous development of diverse needs of users and Web technologies, functionality of SaaS platforms becomes increasingly large, overall performance requirements of SaaS platform are also increasing. Most researches for SaaS platform mainly focused on data storage, multi-tenant areas, quality of service and load balancing, whereas the research specifically for the Web front-end performance optimization is rare, led to front-end performance increasingly became the performance bottleneck of SaaS platform. On the other hand relative to spending a lot of time to improve algorithms and database optimization, front-end performance optimization tends to be able to achieve a multiplier effect. Therefore, providing a appropriate front-end optimization scheme based on the feature of SaaS platform has great practical significance and research space.The paper studied the existing Web performance optimization methods, the optimization methods are classified and summarized. Based on the characteristic of SaaS platform, the paper designed and implemented a optimization scheme from three aspects, which are front-end architecture, resources preloading and HTTP protocol.In terms of front-end architecture, the paper studied JavaScript modular development, asynchronous module load specification, front-end MVC pattern, JavaScript template library, RESTful style server-side interface design, and applied these knowledge to front-end architecture optimization of SaaS platform, and verified the effectiveness of the scheme.In terms of resources preloading technique, the paper analyzed the preloading method compatible with different browsers, and applied the method to users at the most common scenario:user login page and home page. In order to improve the prediction hit ratio, it also designed and implemented a prototype system to predict user behavior, and finally tested its effectiveness.In terms of HTTP protocol optimization, the chapter analyzed the inadequacies of the existing Web based protocol HTTP1.1and studied the SPDY protocol. With the help of the agreement, the HTTP protocol was improved and applied to the existing SaaS platform.