Optimizing Stateful Microservice Migration in Kubernetes with MS2M and Forensic Checkpointing

Abstract: The widespread adoption of microservices architecture in modern software systems has emphasized the need for efficient management of distributed services. While stateless microservices enable straightforward migration, stateful microservices introduce added complexity due to the need to preserve in-memory state during migration. However, most container orchestrators, including Kubernetes, lack native support for live stateful service migration. This paper proposes an optimized migration scheme for stateful services in Kubernetes by integrating the Message-based Stateful Microservice Migration (MS2M) framework with Kubernetes' Forensic Container Checkpointing (FCC) feature. Key enhancements include support for migrating StatefulSet-managed Pods and the introduction of a Threshold-Based Cutoff Mechanism to handle high incoming message rates. Evaluation results demonstrate that MS2M for individual Pods reduces downtime by 96.986% compared to cold migration methods, while the StatefulSet approach provides greater flexibility in managing stateful services. These insights provide practical strategies for optimizing stateful microservice migration in cloud-native environments.