Deadlock detection and recovery for component-based systems

Component-based development has gained a lot of attention in recent years. As a software development paradigm, it enhances reusability and reduces complexity but it also brings new challenges in reliability, especially the deadlock problem. In this paper, we present a dynamic probe (DP) strategy for the deadlock problem of component-based systems (CBS). First a formal semantic model is proposed to abstract the interaction among the components for analyzing the deadlock connections, and then the dynamic probe detection (DPD) algorithm is used to detect the deadlock loops. If deadlock connections are detected then the dynamic probe elimination (DPE) algorithm is used to evaluate the component reliability using several measurement indexes to find a component with reliability lower than the other components, and then replace it. Last, in comparison to related work, the results show that the proposed strategy can achieve both lower processing cost and higher reliability.

► We present a dynamic probe strategy for the deadlock problem of CBS.
► We give a formal semantic model to abstract the interaction among the components.
► We classify the measurement indexes as dynamic and static to evaluate component reliability.
► DPDetection algorithm is given to detect the potential deadlock connections.
► DPElimination algorithm is given to choose the least reliable component to be replaced.