Resource sharing platform architecture for an information product factory

Abstract

Efforts to share resources in collaborative pursuits are hindered by differing data representations, redundant applications, and software incompatibilities. Members of a collaborative effort often span different computational environments and the heterogeneity of contexts disrupts software interoperability. Sharing computational resources has become the focus of many research efforts. Efforts in the 1980s led to the Component Object Model (COM) [Williams, 1998a, 1998b, 1990] and the Common Object Request Broker (CORBA) architectures [Offall et al 1996; OMG, 2003]. In the 1990s both technologies were extended for network support. In recent years, Ian Foster, has phrased the distributed computing problem in terms of sharing computational resources. The grid problem is, "coordinated resource sharing and problem solving in dynamic, multi-institutional virtual organizations", "to support collaborative problem solving in industry, science, and engineering in a data rich environment". [Foster et al, 2001] This work presents a new resource sharing platform architecture for information products that leverages the lessons learned from physical product platforms, the concepts of web services, and grid computing. The platform developed in this thesis integrates the contributions of these three areas into a system that is shown to be more efficient and effective at producing software products.