Handling Fault Tolerance in Edge Computing Using Reactive and Proactive Methods

Abstract

Edge computing provide an effective way to handle the latency oriented Internet of Things (IoT) applications by leveraging the resources of the processing nodes inside the edge network. The device heterogeneity and highly dynamic environment of edge network causes the failures to happen frequently which can result to a service failure or system unavailability. The system availability can be ensured by providing efficient fault tolerance mechanism to IoT applications in general and special for real-time applications as  failures can degrade the performance or stop the processing resulting in problems for the user or a real-time system. This article highlights the different fault tolerance mechanisms including reactive and proactive mechanisms for different processes executing on the edge nodes inside the edge network. And for edge computing we have covered both single edge node execution and distributed application execution. In addition, future research directions are discussed for the researchers working on edge computing