Web Caching

Web Caching

Antonios Danalis (University of Delaware, USA)
DOI: 10.4018/978-1-60566-026-4.ch648
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

The popularity of the World Wide Web has led to an exponential increase of the traffic generated by its users for over a decade. Such a growth, over such a long period of time, would have saturated both the content providers and the network links had Web caching not been efficiently deployed. Web caching can improve the overall performance of the World Wide Web in several ways, depending on the decisions made regarding the deployment of the corresponding caches. By placing caches in strategic positions, the core network traffic can be reduced, the load of a content provider can be scaled down, and the quality of service, as the users perceive it, can be improved. In this article we present an overview of the major design and implementation challenges in Web caching, as well as their solutions.
Chapter Preview
Top

Background

A Web cache can be placed at different positions on the Web and yield different benefits. Namely, it can be placed next to a content provider, in strategic positions inside the network, at the boundaries of local area networks, or inside the host that runs the requesting browser.

Most users tend to access particular pages quite often. An example could be the “home page” that has been set on their browser, or Web sites that contain documentation they are interested in. Furthermore, during browsing, many users visit the same pages multiple times in short periods of time, by using the history of their browsers. To exploit these access patterns, most browsers keep local copies of the frequently or recently accessed pages. This way, the user is served in near zero time, and the network traffic is reduced. Although this technique can have a positive effect on the browsing experience of a single user, it has a minimal effect on the network traffic. This is mostly due to the small disk space used by this type of caching and the lack of sharing between different user caches.

To yield significant results, dedicated computer systems, called proxies, are installed at the edges of local or wide area networks. These systems can achieve significant reduction in the network traffic and improvement in the user perceived quality of service, by filtering and serving the Web requests generated inside the entire network they serve. If a user has defined in the browser’s settings a particular proxy to be used, every time he or she requests a Web page the browser will send this request to the proxy. If the proxy happens to have the page, the user will be served promptly without the original content provider being contacted. If the proxy cannot serve the request, it will fetch the appropriate Web objects (such as the text documents, images, applets) from the original server, and possibly keep a copy for later use. Transparent proxies are a special kind of proxy that end users do not explicitly specify in their browser’s settings. Rather, the gateway of the network identifies Web requests and forwards them to the proxy. This model is more efficient, and easier to administrate, since all the users of the local network will be using the proxy, and they need not know about possible changes concerning the proxy.

To improve the performance of a content provider, a cache can be deployed on the server side. These caches, also called server accelerators, or reverse proxies, are usually deployed in front of clusters of Web servers (Challenger, Iyengar & Dantzig, 1999) and cache the most popular documents in their main memory. Since a few highly popular documents are in most cases responsible for a large percentage of requests, by using a server-side cache, a Web server can decrease its disk I/O load (Abrams et al., 1995; Markatos, 1996; Tatarinov, Rousskov & Soloviev, 1997) and significantly improve the request serving quality of service. Nevertheless, a reverse proxy benefits only a particular content provider and does not reduce the network traffic.

Key Terms in this Chapter

Byte Hit Rate: The ratio of bytes served by the cache over the total number of bytes requested by the clients. BHR can be significantly different from HR in a case where only few, but large files are being served by the cache.

Proxy Cache: A machine dedicated to serving client requests for Web documents. Usually it is installed at the edge of local or wide area networks to provide fast responses to the users and reduce the traffic exiting the network.

HTTP: Hyper-Text Transfer Protocol. The protocol used for most of the communication on the Web, between the clients, the proxies, and the servers.

Transparent Proxy: A proxy that closely cooperates with either a router or a Layer 4 switch, to intercept Web requests while invisible to the users. Other than being invisible it works as a regular proxy cache.

Complete Chapter List

Search this Book:
Reset