作者: Steve Chapin , Paul Talaga
DOI:
关键词:
摘要: The Internet has grown from a static document retrieval system to dynamic medium where users are both consumers and producers of information. Users may experience above-average website latencies due the physical distances information must travel. Because user satisfaction is related website's responsiveness, e-commerce be hindered prevent online businesses reaching their full potential. This dissertation analyzes how temporal relational dependencies in web applications limit ability become distributed. Two contributions made, first showing location data inside datacenter influences system's performance, secondly, that relaxing strict consistency application at fine-grained level can greatly lower for geographically diverse users. Experiments used show when much these optimizations benefit application.