Workflow Optimisation for e-Science Applications David W. Walker, Omer F. Rana, Yan Huang, and Lican Huang School of Computer Science, Cardiff University, 5 The Parade, Roath, Cardiff CF24 3AA, U.K. {david,omer,coral,lican}@cs.cf.ac.uk Abstract. Grid computing is converging on a service-oriented architecture in which applications are composed from geographically dispersed, interacting Web services, and expressed in a workflow description language. The discovery of resources is of key importance to the effective and flexible use of such an architecture. This paper describes work to date in the Workflow Optimisation Services for e-Science Applications (WOSE) project which is investigating techniques for discovering services and optimising the execution of service- composite workflows. Keywords. Grid computing, Web services, workflow, optimisation. 1. Introduction The Grid computing community is now converging on a service-oriented architecture in which applications are composed from geographically dispersed, interacting Web services, and expressed in a workflow description language, typically based on XML. Workflow techniques generally enable a collection of services to be combined dynamically. However, although there is broad consensus on the overall architecture of the Grid there are many unresolved issues that are still active research areas and for which implementations are not publicly available. Grid computing presents many challenges to both middleware developers and application scientists. Amongst these is the discovery of resources to perform a particular task or application, and where multiple resources perform the same function, selecting the optimal one with respect to a set of user-specified criteria. The importance of such optimisation issues is highlighted in service-rich environments in which an application may be able to select from a number of semantically equivalent services that may be characterised by differing performance, cost, and quality of solution. By definition, such optimisation needs to be undertaken in a dynamic environment, within which resource properties can change. Optimisation therefore must utilise a number of heurtistic techniques to enable service selection. This paper presents results from the Workflow Optimisation Services for e-Science Applications (WOSE) project, which is investigating techniques for optimising applications composed of Web services, and developing a prototype system that demonstrates the findings of the research. One of the most challenging aspects of Grid infrastructure development is that there is no central control or authority. Thus, middleware developers can make only rather weak assumptions about third party Web services and the supporting infrastructure of the host environment. It is hoped that standardisation efforts will ameliorate this situation in the future, but standardisation itself is unlikely to resolve all issues. Service discovery is one area that raises many open questions – even deciding whether two given services are equivalent is non-trivial. This paper proposes a software framework for service discovery and selection. Tradeoffs between static and dynamic service discovery are addressed in terms of when tasks are bound to specific services. Interoperability between workflow description languages is also considered. Features of the framework are illustrated with an implementation that supports dynamic invocation of Web services that are coordinated using the Business Process Execution Language (BPEL) workflow description language. 2. Related Work Significant work has already been undertaken in the area of Grid-based workflow systems. A good survey can be found in [3]. The focus of these workflow systems vary – ranging from specialist workflow editors/composition tools, portal technologies to assess the current 27 th Int. Conf. Information Technology Interfaces ITI 2005, June 20-23, 2005, Cavtat, Croatia