Building self-manageable web service processes

点击翻译按钮获取中文摘要

The objective of this project is to study the techniques for building self-manageable Web service processes. Web services are software components which communicate and interact with each other using standard Internet protocols. Software components developed in different programming languages and deployed on different platforms can easily be invoked if they are wrapped as Web services. This is because Internet protocols, such as HTTP and the information format XML used by Web services, are commonly used by many programming languages and on many platforms. Web services are not for human beings to browse, but for software to operate automatically. Web services are a new generation of middleware that have been successfully used for system integrations problems, e.g. connecting heterogeneous data sources. There is a demand to build fully automatic business processes by connecting distributed Web services. Current techniques allow us to build automatic Web service processes in a predefined manner. A Web service process can be defined in a Web service process description language such as BPEL. In such a process, all the Web services are fixed; how they interact with each other is determined at during their design. This process cannot change at run time. If any faults occur, the process simply halts. It would be ideal if the business process could automatically select the best Web services to achieve its goals and, if a Web service were to break down, the business process could switch to other similar Web services and compensate the failure effects. A self-manageable Web service process is a fully automated Web service process with mechanisms for adapting to changes, such as run time execution monitoring, sophisticated way of fault diagnosis and recovery, run time partners selection and changing. As Web services are widely used in e-business applications, scientific computing and software systems, this project will bring great impact.