Assessing Fault Occurrence Likelihood for Service-Oriented Systems

Abstract

Automated identification and recovery of faults are important and challenging issues for service-oriented systems. The process requires monitoring the system's behavior, determining when and why faults occur, and then applying fault prevention/recovery mechanisms to minimize the impact and/or recover from these faults. In this paper, we introduce an approach (defined FOLT) to automate the fault identification process in services-based systems. FOLT calculates the likelihood of fault occurrence at component services' invocation points, using the component's past history, reputation, the time it was invoked, and its relative weight. Experiment results indicate the applicability of our approach.

Integrating Fault-Tolerance Method into Service-Oriented Computing

“…In general, before invoking any service the Service Invocation Module sends the service's information to the History Module which is responsible for calculating the fault likelihood [2]. Based on this value, the Estimation Module determines if the system needs pre-recovery, or it can continue the invocation process.…”

“…In our previous work [2], we defined a fault management approach (Fault Occurrence Likelihood esTimation: FOLT) for SOAs. We assume that component services do not share their execution details with the invoking service (defined as an orchestrator).…”

Automated self-healing framework for service-oriented systems

“…The primary post-recovery action is using WS-BPEL to handle the fault, and if BPEL's built in strategies can not recover from the fault, then as a secondary action, the system re-initiates the planning process. sends the service's information to the History Module which is responsible for calculating the fault likelihood [1]. Based on this value, the Estimation Module determines if the system needs pre-recovery, or it can continue the invocation process.…”

“…However, fault recovery plan generation is challenging due to the lack of capabilities in current systems to adapt themselves at run time to cope with dynamic changes in user requirements and the running levels of QoS attributes [15]. In order to support such dynamic changes, we propose a faulttolerant mechanism which combines our planning strategies based on FOLT calculations [1] and incorporates the exception handling constructs of BPEL. WS-BPEL is an OASIS standard to describe Web service composition.…”

“…In our previous work [1], we defined a fault management approach (Fault Occurrence Likelihood esTimation: FOLT) for SOAs. Once faults are identified, recovery plans need to be created.…”