Abstract
Cloud computing is a cost-effective method of delivering numerous services in Industry 4.0. The demand for dynamic cloud services is rising day by day and, because of this, data transit across the network is extensive. Virtualization is a significant component and the cloud servers might be physical or virtual. Containerized services are essential for reducing data transmission, cost, and time, among other things. Containers are lightweight virtual environments that share the host operating system's kernel. The majority of businesses are transitioning from virtual machines to containers. The major factor affecting the performance is the amount of data transfer over the network. It has a direct impact on the migration time, downtime and cost. In this article, we propose a predictive iterative-dump approach using long short-term memory (LSTM) to anticipate which memory pages will be moved, by limiting data transmission during the iterative phase. In each loop, the pages are shortlisted to be migrated to the destination host based on predictive analysis of memory alterations. Dirty pages will be predicted and discarded using a prediction technique based on the alteration rate. The results show that the suggested technique surpasses existing alternatives in overall migration time and amount of data transmitted. There was a 49.42% decrease in migration time and a 31.0446% reduction in the amount of data transferred during the iterative phase.