Efficient and Scalable Dynamic Graph Replication for Cloud Computing Services

The cloud environment has garnered significant attention due to its crucial role as a supportive framework in computer science and engineering activities. This ever-growing adoption has increased among users, necessitating ongoing efforts to maintain and enhance cloud performance. A widely recognised approach to improving the user experience in cloud environments is data replication. This article proposes a Locality-Aware Dynamic Data Replication Algorithm (LADRE) designed to balance network load and enhance performance. The proposed model identifies the most appropriate data files for replication and selects optimal storage nodes for placing replicas. Simulations conducted using the CloudSim library demonstrate that LADRE improves cloud network load balancing by 5% compared to prior methods. Furthermore, LADRE outperforms existing approaches regarding network response time, storage efficiency, and the frequency of data removal operations.