Web server QoS management by adaptive content delivery

The Internet is undergoing substantial changes from a communication and browsing infrastructure to a medium for conducting business and selling a myriad of emerging services. The World Wide Web provides a uniform and widely-accepted application interface used by these services to reach multitudes of clients. These changes place the web server at the center of a gradually emerging e-service infrastructure with increasing requirements for service quality, reliability, and security guarantees in an unpredictable and highly dynamic environment. Towards that end, we introduce a web server QoS provisioning architecture for performance differentiation among classes of clients, performance isolation among independent services, and capacity planning to provide QoS guarantees on request rate and delivered bandwidth. We present a new approach to web server resource management based on web content adaptation. This approach subsumes traditional admission control based techniques and enhances server performance by selectively adapting content in accordance with both load conditions and QoS requirements. Our QoS management solutions can be implemented either in middleware transparent to the server or by direct modification of the server software. We present experimental data to illustrate the practicality of our approach.