As more and more applications entail interaction among distributed collections of software components, it is becoming increasingly difficult to discern the boundaries between "application", "middleware", "client", and "server". There is an increasing demand for building internet-scale distributed applications, which have the following key characteristics: (a) lack of a single organizational "locus of control" for application development, (b) heterogeneity -- use of a wide variety of languages and systems for each application component, and (c) greater likelihood of system or network failure than in distributed applications interconnected over a local area network.
Despite work on various "infrastructure" standards for Web Services (e.g., SOAP), until recently, there has been little work on programming models directly supporting internet-scale distributed applications. However, in late 2002, the Business Process Execution Language for Web Services standard ("BPEL4WS", or "BPEL"), jointly developed by IBM, Microsoft, and BEA, was released. The name is somewhat misleading -- BPEL is more than a business process specification. It is essentially a high-level programming language, one which addresses some, but not all, of the issues relevant to developing internet-scale distributed applications.
This tutorial will first discuss general issues that arise in building internet-scale applications. Next, we will give a detailed overview of BPEL with examples, showing how BPEL addresses many of these issues. The overview will cover processes, service links activities, containers, scopes, fault handlers, and exception handlers. Finally, we will cover a number of other technologies and standards related to programming internet-scale distributed applications, and discuss their relationship to BPEL.
Prerequisites: Basic knowledge of object-oriented languages, web services, and parallel and distributed computing.
Dr. Vivek Sarkar is Senior Manager of the Programming Technologies department at the IBM T. J. Watson Research Center. His research interests are in the areas of static and dynamic program analyses and their applications to optimization, parallelization, and programming tools. The projects under way in his department include component verification (Canvas), analysis and optimization of web service and database applications (DOMO), refactoring analysis and transformation (Gnosis), adaptive optimization and memory wall (Jikes RVM), data model extraction from legacy applications (Mastery), and dynamic analysis and datarace detection (Shrike). Vivek obtained his Ph.D. from Stanford University in 1987, and has been a member of the IBM Academy of Technology since 1995.
John Field is a Research Staff Member and manager of the Advanced Programming Tools group at IBM's T.J. Watson Research Center. His research interests include static analysis and verification of programs, tools for program understanding and transformation, program slicing, program semantics and logics, and programming models for distributed systems. In addition to publishing his work in a wide variety of conferences and journals, he has contributed to the development of several IBM products. He received a Ph.D. from Cornell University in 1991.