摘要: Achieving high dependability and fault-tolerance in service-oriented architecture (SOA) is an open problem. Exception handling is one of the powerful means for improving the quality of SOA. The paper discusses the results of exper... 展开 Achieving high dependability and fault-tolerance in service-oriented architecture (SOA) is an open problem. Exception handling is one of the powerful means for improving the quality of SOA. The paper discusses the results of experimental analysis of the SOA-specific exceptions and factors affecting availability and faulttolerance of Web Services, implemented using two development kits: JAX-RPC implementation at Sun Microsystems and IBM WebSphere Software Developer Kit for Web Services. We specifically focus on the results of exception propagation and performance analysis. Finally, applications of different error recovery strategies including backward, forward and enhanced forward error recovery, in the context of SOA are briefly discussed. 收起