Green computing for improving the sustainability of data centers: Optimized VM allocation with decentralized peer-to-peer nodes

Abstract

The amplified concern in cloud and fog computing calls for a stimulus in the demands on the strategies that allocate virtual machines to amplify energy efficiency to alleviate high energy utilization and the environmental impacts that arise from it. This research presents a green computing approach that employs decentralized, peer-to-peer fog nodes in dynamic VM allocation within the fog and cloud ecosystems. Unlike traditional central allocation schemes, the proposed model allows for autonomous management, distribution, and sharing of workload at the level of a fog node, depending upon local capacity, real-time demands, and determined energy efficiency. P2P collaboration at geographically distributed fog nodes evokes optimal resource usage, minimal latency, energy costs, and carbon footprint. An energy-aware allocation algorithm is developed that integrates real-time workload prediction, power consumption metrics, and renewable energy availability across the boundaries of fog and cloud environments to enhance sustainability. Experimental results demonstrate that the decentralized P2P framework not only diminishes power utilization but also improves the response times for services and minimizes the overall environmental footprint of fog/cloud operations.