Article Published In Vol.9 (Sept-Oct 2021)

Optimizing Cloud Resource Management via Dynamic Auto-Scaling in E-Commerce Applications: Enhancing Load Balancing and Performance Efficiency

Author : Rohith Reddy Mandala and Thanjaivadivel M

Resource utilization in e-commerce services is a critical factor in maintaining performance, scalability, and cost efficiency within the evolving landscape of cloud computing. Effective resource allocation ensures that systems can handle fluctuating traffic loads, minimize downtime, and optimize operational costs while providing a seamless user experience. Resource management systems applied to modern cloud environments also generate resource waste or overloading as a result of over-provisioning, insufficiency of resources, or inefficient scaling, making operational efficiency and service-level agreement (SLA) penalties complex. These flaws require the implementation of a more data-driven, responsive resource distribution mechanism. Our research proposes a dynamic auto-scaling model integrated with ARIMA-based predictive scaling, which enhances load balancing and performance efficiency. Using predictive models, the system automatically allocates resources ahead of workload predictions, reducing costs by up to 30% and system uptime by up to 99.5%. These inefficiencies in scaling challenges are met by our model by providing optimal usage of resources with predictive scaling, ensuring reduced waste and improved responsiveness. Compared to conventional scaling, the model reflects an impressive boost in efficiency by providing 30% cost reduction, 20% improved resource usage, and 4.5% improvement in uptime. The findings confirm that our method performs better than conventional scaling methods regarding resource usage and cost savings. This research will add to the design of a smart and scalable cloud resource management solution for e-commerce websites that will provide the best performance, lower costs, and improved user experience.

Keywords: Dynamic Auto-Scaling, Load Balancing, ARIMA Model, Predictive Scaling, Resource Allocation, Cost Optimization, Scalability.