AI-Driven Predictive Autoscaling in Kubernetes : Reinforcement Learning for Proactive Resource Optimization in Cloud-Native Environments
DOI:
https://doi.org/10.32628/CSEIT22548Keywords:
Kubernetes autoscaling, reinforcement learning, cloud cost optimization, predictive analytics, AWS EKS, KarpenterAbstract
We propose a reinforcement learning-based autoscaling algorithm integrated with Karpenter on AWS EKS. Unlike threshold-based scaling, our method anticipates workload surges by analyzing historical patterns using predictive analytics, thereby reducing cloud spend and improving service availability. Simulations and real deployment benchmarks from Rialtic Inc. validate the cost efficiency and reliability of this method. The proposed system achieves 34% reduction in cloud infrastructure costs while maintaining 99.7% service availability and reducing cold start latencies by 67%. Through Q-learning optimization and temporal pattern recognition, the system demonstrates superior performance compared to traditional Horizontal Pod Autoscaler (HPA) and Vertical Pod Autoscaler (VPA) mechanisms.
References
- Chen, L., Wang, S., & Liu, Y. (2019). Analysis and optimization of Kubernetes horizontal pod autoscaler for cloud-native applications. Proceedings of the International Conference on Cloud Computing and Services Science, 245-256.
- Delimitrou, C., & Kozyrakis, C. (2018). Quasar: Resource-efficient and QoS-aware cluster management for heterogeneous workloads. ACM Transactions on Computer Systems, 36(4), 1-32.
- Santhosh Kumar Pendyala, Satyanarayana Murthy Polisetty, Sushil Prabhu Prabhakaran. Advancing Healthcare Interoperability Through Cloud-Based Data Analytics: Implementing FHIR Solutions on AWS. International Journal of Research in Computer Applications and Information Technology (IJRCAIT), 5(1),2022, pp. 13-20. https://iaeme.com/Home/issue/IJRCAIT?Volume=5&Issue=1
- Kumar, A., Singh, R., & Patel, N. (2018). Cost-aware workload scheduling in Kubernetes clusters: A multi-objective optimization approach. Journal of Cloud Computing Advances, Systems and Applications, 7(1), 15-28.
- Liu, X., & Zhang, H. (2020). LSTM-based predictive scaling for containerized applications in cloud environments. IEEE Transactions on Cloud Computing, 8(3), 892-904.
- Sushil Prabhu Prabhakaran, Satyanarayana Murthy Polisetty, Santhosh Kumar Pendyala. Building a Unified and Scalable Data Ecosystem: AI-DrivenSolution Architecture for Cloud Data Analytics. International Journal of Computer Engineering and Technology (IJCET), 13(3), 2022, pp. 137-153. https://iaeme.com/Home/issue/IJCET?Volume=13&Issue=3
- Patel, M., Johnson, K., & Thompson, R. (2020). Q-learning optimization for virtual machine placement in heterogeneous cloud environments. International Journal of Cloud Computing and Services Architecture, 10(2), 34-47.
- Williams, D., & Thompson, A. (2019). Cost optimization strategies for containerized workloads: An empirical study of Kubernetes deployments. Cloud Computing and Services Science Communications, 425, 156-170.
- Zhou, P., Chen, M., & Davis, J. (2017). Reinforcement learning for adaptive resource allocation in cloud computing: A comprehensive survey. ACM Computing Surveys, 50(6), 1-38.
Downloads
Published
Issue
Section
License
Copyright (c) IJSRCSEIT
This work is licensed under a Creative Commons Attribution 4.0 International License.