Zhu Jun, Guo Changguo, Wu Quanyuan. A Runtime Monitoring Web Services Interaction Behaviors Method Based on CPN[J]. Journal of Computer Research and Development, 2011, 48(12): 2277-2289.
Citation:
Zhu Jun, Guo Changguo, Wu Quanyuan. A Runtime Monitoring Web Services Interaction Behaviors Method Based on CPN[J]. Journal of Computer Research and Development, 2011, 48(12): 2277-2289.
Zhu Jun, Guo Changguo, Wu Quanyuan. A Runtime Monitoring Web Services Interaction Behaviors Method Based on CPN[J]. Journal of Computer Research and Development, 2011, 48(12): 2277-2289.
Citation:
Zhu Jun, Guo Changguo, Wu Quanyuan. A Runtime Monitoring Web Services Interaction Behaviors Method Based on CPN[J]. Journal of Computer Research and Development, 2011, 48(12): 2277-2289.
1(The 63rd Research Institute of PLA General Staff Headquarters, Nanjing 210007) 2(College of Computer, National University of Defense Technology, Changsha 410073) 3(China Electric Equipment and Systems Engineering Ltd., Beijing 100039)
BPEL (business process execution language) is one of the dominant ways to specify service interactions between different Web services to implement much more complex functions. Since it is a kind of description language for Web services composition, BPEL has difficulty in dealing with behavioral properties of service compositions. Usually, well-defined interaction protocols may be violated by clients and other abnormal partnership Web services, and it leads the service composition processes to inconsistent states and exceptions. As a result, we propose to tackle the conformance problem between interactions of Web services and its description specification by using an automatically-generated runtime monitor from the BPEL description. Firstly, a formal representation model based on colored Petri net (CPN) is introduced to extract the service interaction behaviors from its description. The pattern mapping rules from BPEL description to colored Petri net model and related embedding, reduction and composition rules are also provided. Then a runtime monitor is generated, which will capture service interaction behaviors fromto the service composition processes and detect inappropriate use of the interaction protocol. Several typical service composition samples are adopted as case study. Finally, full evaluations show that this runtime monitoring mechanism costs low overhead and has good performance and efficiency.