Web服务组合的行为推断诊断方法

随着Web服务以及Web服务组合应用软件在分布式网络中的广泛应用,Web服务的规模和复杂性也在不断地增加,这使得服务在运行过程中可能产生各种故障,因此对服务系统进行及时的故障诊断与排除越来越重要.为了解决在故障诊断中系统模型不完备和历史数据中存在噪音数据这一实际问题,提出一种基于服务行为模型的行为推断诊断方法.该方法通过加权方式结合多种诊断信息构建服务行为模型,应用隐马尔科夫模型中的解码思想推断出与异常执行序列最匹配的正常执行序列,并与观察序列进行比较,从而发现差异定位服务故障.实验表明,该方法应用包含不同噪音比例的诊断信息进行诊断,其诊断准确性均高于传统的服务故障诊断方法.

Diagnosis Method of Behavior Inference in Web Service Composition

With the wide applications of Web services and composite services in the distributed network,the size and complexity of Web service are increasing continuously.These could cause various faults of service system during the running.In building high-reliable service applications,one of the critical challenges is how to localize faulty service quickly and exactly and help service engine restore the normal process as soon as possible.To perfect the diagnosis mo-del and minimize the impact of noise data on diagnosis accuracy,we presented a service behavior model-based diagnosis method of behavior inference.The method models hidden markov model by combining historical data into service process definition.On the basis of using the decoding algorithms in HMM,the method is able to infer a correct execution trace which has the maximum likelihood with the exception execution trace and localize the service faults by comparing the differences between them.The experimental results show that the method is effective and robust to various noises in diagnosing the faults of Web services.