Srikanth Nimmagadda

Abstract

All Ensuring business continuity in modern enterprises requires storage systems capable of withstanding hardware failures, network disruptions, and site-wide outages without compromising data integrity or availability. Distributed storage architectures provide high scalability and redundancy; however, the choice of replication strategy is critical for achieving fault tolerance. This paper investigates and compares synchronous, asynchronous, and hybrid replication strategies in distributed storage systems, emphasizing their suitability for mission-critical workloads. Using a Linux-based experimental testbed with both SAN and object storage backends, we simulate failure scenarios and evaluate metrics such as Recovery Point Objective (RPO), Recovery Time Objective (RTO), latency overhead, and bandwidth consumption. The study leverages open-source platforms like Ceph and MinIO, automated deployment through Ansible, and real-time monitoring via Grafana. Results highlight key trade-offs betw