Designing a Cost-efficient Model for Replicating Data in Cloud Storage
Muhammad Imran, Institute of Computer Sciences and Information Technology (ICS/IT), The University of Agriculture Peshawar, Pakistan.
Syed Awais Ali Shah, Institute of Computer Sciences and Information Technology (ICS/IT), The University of Agriculture Peshawar, Pakistan.
Arshad Khan, Institute of Computer Sciences and Information Technology (ICS/IT), The University of Agriculture Peshawar, Pakistan.
Corresponding Author:
Muhammad Imran (mimran@aup.edu.pk)
Abstract:
Cloud storage systems use replication technologies to enhance data availability and decrease data access latency by providing users with multiple copies of the same data. However, the cost of redundant storage and maintaining the consistency of replicas place a significant burden on cloud subscribers. Current approaches to address the issue focus on improving system performance but ignore the costs associated with determining the number of replicas to create and their storage locations. To address this challenge, a cost-effective dynamic replication algorithm is proposed that reduces the number of replicas on nodes while maintaining availability, ensuring efficient distribution of resources. The proposed model is tested and validated with 25% and 75% threshold values to determine its cost-effectiveness. The comparison is done with Microsoft's Azure cloud computing platform, and the findings reveal that the proposed scheme is more cost-effective than the Azure system when triple replication is used.
Keywords:
Cloud Computing; Data Replication; Cost-Efficient