Automated Disaster Recovery as a Service: A Continuous Resilience Approach for Cloud Environments

In today's rapidly evolving market, business continuity has become a critical determinant of organizational success. As online services and the digital economy continue to expand, ensuring the uninterrupted operation of IT services has grown increasingly important. Disaster Recovery as a Service (DRaaS) provides customers with disaster recovery capabilities on a payas- you-go basis. However, existing DRaaS models often depend on manual development and implementation processes, requiring frequent updates to maintain their effectiveness. A solution was presented featuring a framework that integrated the disaster recovery process directly into the deployment stage, providing full control over both deployment and recovery planning without requiring human intervention. In addition, the framework incorporates a Long-Short-Term Memory (LSTM)-based algorithm designed to predict future resource utilization. This predictive capability enhances decisionmaking during incidents by offering proactive insights, which in turn facilitates more efficient resource allocation and improved response strategies. The framework can be seamlessly embedded into DevOps workflows. To evaluate its effectiveness, an environment simulating real-world business operations and production failures was established, and experimental results confirmed the efficacy of the proposed solution. Additional components are also proposed to enhance the framework's efficiency and adaptability when handling complex scenarios.