Sedna: A BPEL-Based Environment for Visual Scientific Workflow Modeling

Scientific Grid computing environments are increasingly adopting the Open Grid Services Architecture (OGSA), which is a service-oriented architecture for Grids. With the proliferation of OGSA, Grids effectively consist of a collection of Grid services, Web services with certain extensions providing additional support for state and life cycle management. Hence, the need arises for some means of composing these basic services into larger workflows in order to, for example, express a scientific experiment. There are several approaches for composing Web services into larger work-flows, most of which, at least for the composition of scientific workflows, are based on custom-made workflow languages and corresponding enactment environments. Another approach, which we have taken in our work, is to use an industry standard for the orchestration of Web services, such as the Business Process Execution Language (BPEL) [82]. BPEL, which has been introduced by IBM and Microsoft, makes a number of benefits available to scientific Grid computing. The host of commercial providers supporting BPEL means that there are industrial-strength enact-ment environments and middleware technologies available that exhibit a level of scalability and reliability that a research prototype could not match. The multitude of providers supporting BPEL creates a market, which means that it is a live standard with ongoing efforts to develop new features. Furthermore , BPEL could serve as a standard representation for scientific workflows and hence aid reproducibility. Finally, as a programming language that fo-cuses on high-level state transitions, it could enable computational scientists to compose scientific workflows themselves, relieving them of a dependence on software engineers. In our work, we have been investigating the applicability of BPEL for the expression of scientific workflows. We have established in a companion paper that freely available BPEL enactment environments satisfy the scalability and