Efficient Container Migration for Cloud Applications: A Secure and Cost-Effective Pre-Copy Strategy

Container migration is a critical capability in cloud computing, enabling dynamic resource allocation and maintaining high availability for applications. This study proposes an optimized container migration technique to enhance migration efficiency by minimizing downtime and reducing data transfer overhead. Unlike traditional methods—Cold, Pre-copy, Post-copy, and Hybrid—our approach utilizes a predictive algorithm to identify and transfer only the essential memory pages during the pre-dump phase. Experimental evaluations show that the proposed method achieves substantial improvements, with a total migration time of 40 seconds compared to 64–95.67 seconds in other techniques and a downtime reduction to 18 seconds, significantly lower than traditional methods. Additionally, our technique minimized data transfer to 5700 bytes, optimizing network usage in large-scale cloud environments. These results confirm that the proposed approach is well-suited for high-demand cloud applications requiring minimal disruption and efficient resource utilization. Future work may extend this method by incorporating machine learning models to adapt to varying workloads and enhance predictive accuracy dynamically.