The web sites that fuel e-commerce rely increasingly on applications that generate dynamic content to provide visitors with personalized, interactive experiences. As Internet traffic continues to grow and web sites become more complex, performance and scalability are fast becoming major issues for many organizations. At the same time, today’s e-commerce companies are looking for more cost-efficient ways to balance the need to meet growing customer demands against the pressure to control IT costs.
These issues are addressed in recent research and a pilot study conducted by Kaushik Dutta and Debra VanderMeer, both assistant professors in the Department of Decision Sciences and Information Systems in the College of Business Administration, who collaborated with three colleagues to publish their findings in an article titled “Proxy-Based Acceleration of Dynamically-Generated Content on the World Wide Web: An Approach and Implementation.” The work appeared in ACM Transactions on Data Systems, published by the Association of Computer Machinery—an organization that delivers resources to advance computing as a science and a profession.
Caching 101 and beyond.
Simply put, caching is a way to store bits of data to be reused, eliminating the need for repeated computations each time someone requests that data. Cached data can be stored on an Internet browser, a web server, or on a node somewhere in between. In this context, proxy-based caching refers to a computer machine—or “proxy”—that serves cached content in place of the site of origin.
“It’s easy to cache a static image or text,” VanderMeer said. “It’s much more difficult to cache personalized or dynamically-generated content. That’s where our research comes in. Our dynamic proxy caching technique is novel in that it enables proxy-based caching in a fully-distributed mode, allowing both the content and the layout to be dynamic. In less technical terms, this means that we devised a way to cache part of a web page instead of caching the whole web page seen in the browser.”
The research team’s subsequent analysis of the performance of this approach indicated that it is capable of providing significant reductions in bandwidth and response times—with the ability to scale with respect to the number of cached fragments as well as with the number of proxies.
Pilot demonstrates results with real business value.
Dutta, VanderMeer, and their colleagues were able to run a pilot deployment of their proposed caching technique at a major financial institution.
“The results indicated that our technique is capable of providing up to three times the reductions in bandwidth and response times in real-world dynamic web applications when compared to existing caching solutions,” Dutta said. “This result provides a viable solution for delivering more processing power to support more users—without having to purchase more hardware and software.”
Such a technology solution in turn translates into real total cost benefits that resonate equally well with both the IT and business sides of any e-commerce company.