Abstract
Cluster-based servers are proven to be the best solution to build high-performance, scalable, and reliable Internet Web servers. A number of researches try to enable the dispatcher in cluster-based web servers to route the users requests based on higher layer information, so more detail information such as URLs in the content of packets will be used for dispatching decisions. Hashing functions and tree structure are used to achieve the goal of URL lookup, but they will cause the problem of collision and slow speed. This thesis presents a fast URL lookup mechanism that uses CAM as the basic hardware architecture. Our scheme not only supports exact matching of URLs, but also provides URL prefix matching to provide greater expansibility and flexibility. The proposed scheme takes constant time to lookup an URL and provides high scalability. With the support of our fast URL lookup engine, the performance of content-aware switches and other applications requiring content information can be largely improved.