A Fault-Tolerant Workflow Composition and Deployment Automation IoT Framework in a Multicloud Edge Environment

Abstract

With the increasing popularity of IoT application, e.g., smart home, smart manufacturing, the importance of underlying system availability, safety, reliability, and maintainability become crucial in the development processes. IoT applications are expected to continuously provide reliable services and features, which put the fault-tolerance mechanism as a priority. Current IoT fault-tolerant systems are designed to overcome any faults caused by human activities and physical errors to preserve the correct IoT workflow execution. However, addressing fault-tolerant interaction in multicloud edge environment and failed service deployment automation remains challenging. This article proposes a novel fault-tolerant model that offers the self-detection and automatic recovery of faults to increase IoT applications reliability and to address the infrastructure level failure in the heterogeneous IoT environments. Based on the proposed model, we implement a fault-tolerant workflow composition and deployment automation system, which is empowered by a layered architecture and a time-dependent failure model. The efficiency and effectiveness of the proposed system are validated and evaluated with a real-world IoT application.