A Dynamic Lifecycle-Aware Disaster Recovery Framework for Stateful IoT Applications in Edge Environments

초록

Internet of Things (IoT) systems are increasingly deployed across remote and distributed environments, enabling edge computing capabilities that span multiple lifecycle stages such as deployment, configuration, runtime, failure recovery, and updates. These systems are often decomposed for scalability and adaptability but require integrated coordination across infrastructure, applications, and devices. In disaster recovery scenarios, maintaining stateful continuity across these layers is essential yet remains challenging. Existing migration techniques often incur high overhead, focus narrowly on runtime preservation without lifecycle context, disrupt system modularity, and depend on manual or one-time recovery actions. This paper presents a systematic, GitOps-driven recovery framework for stateful IoT systems, unifying infrastructure, services, and device management under a declarative lifecycle model. The framework automates stateful migration through lifecycle-aware orchestration that preserves configuration versions, topology structure, and orchestration semantics while abstracting infrastructure heterogeneity. Furthermore, it introduces device-binding-aware recovery, treating physical connectivity of device-to-system as first-class declarative resources that can be consistently recovered. We implement and evaluate the framework across diverse IoT workloads and edge platforms, demonstrating significant improvements in recovery time, state consistency, and management overhead compared to conventional migration and failover solutions.

키워드