A Fault Detection Mechanism for SOA-Based Applications Based on Gauss Distribution

Service-oriented architecture (SOA) is an ideal solution to build application system with low cost and high efficiency, but fault detection is not supported in most SOA-based applications. Based on Gauss distribution, a fault detection mechanism for SOA-based applications is proposed. The fault in SOA can be detected through comparing the calculated confidence interval with the predefined parameters at runtime according to the descriptor. Based on the fault detection algorithm, the reference service model is improved to support the proposed algorithm through adding some suitable components.