Publications

Abstract : As Internet services become more popular and pervasive, a serious problem that arises is managing the performance of services under intense overload. Models and mechanisms proposed by researchers till date either stop answering requests or give answers based on incomplete/stale data, specifically when Web servers become saturated. The paper provides two contributions to the study of improving the performance of Web servers in terms of support for massive concurrency. Firstly it outlines the factors necessary to build scalable Web servers and the research contributions made to these. Secondly it illustrates how these factors can be combined and applied to obtain architecture for highly concurrent Internet services. The proposed scheme addresses the issues of overload and staleness effectively and ensures that more number of legitimate requests is answered with complete data. It is shown to improve Web QoS for commercial Web servers through control mechanisms that offer adaptive load shedding,improved availability, and secure transactions using a multi layered filtering scheme. Our experimental results show that the proposed architectural framework provides a simpler programming model that achieves superior performance.