The features of heterogeneous, distributed and dynamic services have resulted in more unpredictable failure in service-oriented computing, which decrease the reliability of composite services composed of individual services. Study on how to guarantee the reliability of composite services in a distributed, dynamic, and complex environment becomes more and more important. In this paper, a method based on redundant mechanism as well as a management mechanism is proposed to improve the reliability of the composite services. Our contribution in this paper is threefold. First, VTS, defined as a group of redundant services which is virtualized to a normal service, is designed to manage the redundant services apparently and inner management mechanism of VTS based on event-rule-strategy is put forward to help VTS adapt to the QoS adjustment of the running services. Second, a feedback control-based framework of service running platform is designed. Last but not least, based on results above, RSSA, a selection approach of redundant services is designed. All above works' optimal effect on the reliability of composite services is demonstrated through the experiments.
展开▼
