As the Web is becoming a medium widely used as a preferential channel for critical information exchange, business, and e-commerce, it is necessary to enable differentiated service mechanisms not only at the network but also at the Web server level. In this paper, we propose the concept of Quality of Web Services (QoWS), which is inspired by the basic principles of network QoS, while looking at the server components of the Web system. In particular, we analyze how QoWS principles can be realized in a Web site hosted on a Web-server cluster that is, an architecture composed by multiple Web servers locally distributed and a single front-end node, called a Web switch. We propose a new centralized policy, namely DynamicPartitioning, which satisfies through dynamic server partition all basic QoS principles for a Web switch working at application level. We compare it against other proposed classes of policies which implement part or all of basic QoS principles. We demonstrate through a large set of simulation experiments under a realistic workload model that DynamicPartitioning always achieves superior performance for the high service class, at the price of some penalty for low service classes.
Cardellini, V., Casalicchio, E., Colajanni, M., Mambelli, M. (2001). Web switch support for differentiated services. PERFORMANCE EVALUATION REVIEW, 29(2), 14-19 [10.1145/572317.572320].
Web switch support for differentiated services
CARDELLINI, VALERIA;CASALICCHIO, EMILIANO;
2001-09-01
Abstract
As the Web is becoming a medium widely used as a preferential channel for critical information exchange, business, and e-commerce, it is necessary to enable differentiated service mechanisms not only at the network but also at the Web server level. In this paper, we propose the concept of Quality of Web Services (QoWS), which is inspired by the basic principles of network QoS, while looking at the server components of the Web system. In particular, we analyze how QoWS principles can be realized in a Web site hosted on a Web-server cluster that is, an architecture composed by multiple Web servers locally distributed and a single front-end node, called a Web switch. We propose a new centralized policy, namely DynamicPartitioning, which satisfies through dynamic server partition all basic QoS principles for a Web switch working at application level. We compare it against other proposed classes of policies which implement part or all of basic QoS principles. We demonstrate through a large set of simulation experiments under a realistic workload model that DynamicPartitioning always achieves superior performance for the high service class, at the price of some penalty for low service classes.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.