Comparative Analysis of Distributed Storage Systems: Architectural Design, Performance, and Cost Trade-offs in Modern Cloud Environments

With the exponential growth of data and increasing need for remote access, distributed storage systems (DSSes) have emerged as a foundational technology for efficient data management and scalability. Traditional storage models are challenged by the requirements of Internet-based systems, where reliability, scalability, and cost-efficiency are essential. This study explores the transition from proprietary hardware-based sys- tems to cloud-based storage solutions and the pressures driving organizations toward Storage as a Service (STaaS) models. Emphasizing the cost-saving and scalability benefits of using commodity hardware, this paper examines three open-source distributed storage systems—Tahoe Least-Authority Filesystem (Tahoe-LAFS), Ceph, and Lustre—each representing a unique architectural approach to data distribution and management. Our research evaluates these systems based on key metrics, including scalability, availability, durability, performance, dy- namic operation, and cost-effectiveness. Through analysis and implementation of these DSSes in a networked environment, we aim to understand the design trade-offs and performance characteristics inherent in each system, providing insights into the best practices for organizations transitioning to modern, scalable, and resilient storage architectures.