Leveraging Machine Learning for Automated Thread Dump Analysis and Performance Tuning in Enterprise Java Applications
Abstract
Abstract
Thread dumps are a critical diagnostic tool for identifying performance bottlenecks, deadlocks, and resource contention in enterprise Java applications. However, manual analysis of thread dumps is time-consuming and error-prone, especially in large-scale, distributed environments. This paper presents a machine learning-based framework for automated thread dump analysis and performance tuning. By leveraging supervised and unsupervised learning techniques, the system identifies recurring patterns, classifies thread states, detects anomalies, and correlates them with application metrics to suggest actionable optimizations. The framework is evaluated across multiple real-world enterprise Java workloads, demonstrating improved accuracy in issue detection and significant reductions in mean time to resolution (MTTR). The proposed approach enables proactive performance tuning, enhances system reliability, and reduces dependency on expert intervention, paving the way for intelligent, self-healing Java application environments.
References
Jain, J., Modake, R., Khunger, A., & dnyandev Jagdale, A. CLOUD-NATIVE SECURITY FRAMEWORK: USING MACHINE LEARNING TO IMPLEMENT SELECTIVE MFA IN MODERN BANKING PLATFORMS , 2019.
Gregg, B. (2013). Systems performance: Enterprise and the cloud. Prentice Hall.
Gunther, N. J. (2010). Guerrilla capacity planning: A tactical approach to planning for highly scalable applications and services. Springer.
Khokhar, R., & Soni, D. (2017). Machine learning techniques for software performance analysis: A survey. In Proceedings of the International Conference on Intelligent Communication and Computational Techniques (pp. 33–39).
Farjami, Y., Zulkernine, M., & Martin, P. (2018). Automated classification of performance issues in Java applications using machine learning. In Proceedings of the IEEE International Conference on Software Quality, Reliability and Security (pp. 224–229).
Shah, A., & Kumar, A. (2016). Analyzing Java thread dumps using visualization techniques. In Proceedings of the International Conference on Computing, Communication and Automation (pp. 60–65).
