An Empirical Evaluation Of Web Services Architectures From An Economic Perspective

The evaluation of the two principal styles of web services architecture: Representational State Transfer (REST) and Service Oriented Architecture (SOA) has conventionally focused on the extent to which each architectural style induces desirable architectural properties. The increasing prevalence of enterprise web services and the commercialization of publicly available web services has driven increased interest in the association of Quality of Service (QoS) with the provision of services. Accordingly, the evaluation of web services architectures must consider the consequences of each architecture in terms of Quality of Service (QoS), such as performance and availability, and in terms of profit yielded by a provider offering service provision contracts contingent on this QoS. We introduce a utility model to facilitate the formation of optimal service contracts under the uncertainty in QoS experienced using a given architecture. We then introduce the Amazon Simple Storage Service (S3) as a pertinent example service, and present an architecture for contract-based service provision along with an implementation of the architecture using S3. Empirical results obtained from the implementation are presented to illustrate the consequences of a chosen architecture on QoS and therefore on the profit yielded by the provider from service provision contracts. This demonstrates the applicability of an empirical evaluation of web services architectures from an economic perspective. © 2009 University of Newcastle upon Tyne. Printed and published by the University of Newcastle upon Tyne, Computing Science, Claremont Tower, Claremont Road, Newcastle upon Tyne, NE1 7RU, England. Bibliographical details SMITH, C., VAN MOORSEL, A. An Empirical Evaluation Of Web Services Architectures From An Economic Perspective [By] C. Smith, A. van Moorsel. Newcastle upon Tyne: University of Newcastle upon Tyne: Computing Science, 2009. (University of Newcastle upon Tyne, Computing Science, Technical Report Series, No. CS-TR-1148)