Title :
A Fault Detection Mechanism for Fault-Tolerant SOA-Based Applications
Author :
Chen, Hao-Peng ; Wang, Zhi-Yong
Author_Institution :
Shanghai Jiao Tong Univ., Shanghai
Abstract :
Fault tolerance is an important capability for SOA-based applications, since it ensures the dynamic composition of services and improves the dependability of SOA-based applications. Fault detection is the first step of fault detection, so this paper focuses on fault detection, and puts forward a fault detection mechanism, which is based on the theories of artificial neural network and probability change point analysis rather than static service description, to detect the services that fail to satisfy performance requirements at runtime. This paper also gives reference model of fault-tolerance control center of enterprise services bus.
Keywords :
neural nets; probability; software architecture; software fault tolerance; artificial neural network; fault detection mechanism; fault-tolerant SOA-based application; probability change point analysis; service-oriented architecture; Application software; Artificial neural networks; Computer architecture; Delay; Fault detection; Fault tolerance; Machine learning; Performance analysis; Runtime; Service oriented architecture; Artificial neural network; Fault detection; Performance requirement; Probability change point analysis; Soa;
Conference_Titel :
Machine Learning and Cybernetics, 2007 International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-0973-0
Electronic_ISBN :
978-1-4244-0973-0
DOI :
10.1109/ICMLC.2007.4370804