Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{196095,
        author = {GOUTHAM N and DHEERAJ ROHITH J and JEEVAN PRASAD S and TAMILSELVI B},
        title = {Predictive Failure Detection for Cloud and Infrastructure Using LSTM},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {3106-3112},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=196095},
        abstract = {The stability of server infrastructure is paramount in the modern digital economy, where downtime results in significant financial loss and reputational damage. Traditional reactive maintenance models, which rely on static thresholds and manual intervention post-failure, are increasingly insufficient for complex, cloud-native environments. This paper presents the Predictive Server Management System (PSMS), a robust machine learning framework designed to anticipate system failures before they disrupt operations. Utilizing a Random Forest ensemble classifier, the system analyzes real-time telemetry—including CPU load, memory variance, and network latency—to detect non-linear patterns indicative of impending anomalies. We detail the end-to-end architecture, from synthetic data generation and sliding-window feature engineering to the deployment of a low-latency inference engine. Extensive empirical evaluation demonstrates that our approach achieves an accuracy of 94.5% and an F1-score of 92.0%, significantly outperforming baseline Logistic Regression and Support Vector Machine (SVM) models. Furthermore, we explore the socio-economic implications of this technology, emphasizing its role in sustainable” Green IT” by optimizing resource usage and reducing energy-intensive system reboots. The study concludes with a roadmap for integrating this system into containerized Kubernetes environments.},
        keywords = {Predictive Maintenance, Server Management, Random Forest, Anomaly Detection, AIOps, Machine Learning, Feature Engineering, Reliability Engineering, Smart Infrastructure.},
        month = {April},
        }