Security-Enhanced QoS-Aware Autoscaling of Kubernetes Pods Using Horizontal Pod Autoscaler (HPA)
Journal Title: Journal of Intelligent Management Decision - Year 2024, Vol 3, Issue 3
Abstract
Container-based virtualization has emerged as a leading alternative to traditional cloud-based architectures due to its lower overhead, enhanced scalability, and adaptability. Kubernetes, one of the most widely adopted open-source container orchestration platforms, facilitates dynamic resource allocation through the Horizontal Pod Autoscaler (HPA). This auto-scaling mechanism enables efficient deployment and management of microservices, allowing for rapid development of complex SaaS applications. However, recent studies have identified several vulnerabilities in auto-scaling systems, including brute force attacks, Denial-of-Service (DoS) attacks, and YOYO attacks, which have led to significant performance degradation and unexpected downtimes. In response to these challenges, a novel approach is proposed to ensure uninterrupted deployment and enhanced resilience against such attacks. By leveraging Helm for deployment automation, Prometheus for metrics collection, and Grafana for real-time monitoring and visualisation, this framework improves the Quality of Service (QoS) in Kubernetes clusters. A primary focus is placed on achieving optimal resource utilisation while meeting Service Level Objectives (SLOs). The proposed architecture dynamically scales workloads in response to fluctuating demands and strengthens security against autoscaling-specific attacks. An on-premises implementation using Kubernetes and Docker containers demonstrates the feasibility of this approach by mitigating performance bottlenecks and preventing downtime. The contribution of this research lies in the ability to enhance system robustness and maintain service reliability under malicious conditions without compromising resource efficiency. This methodology ensures seamless scalability and secure operations, making it suitable for enterprise-level microservices and cloud-native applications.
Authors and Affiliations
"Vani Rajasekar, Muzafer Saračević, Darjan Karabašević, Dragiša Stanujkić, Amor Hasić, Melisa Azizović, Srivarshan Thirumalai"
Keywords
Related Articles
- EP ID EP752390
- DOI https://doi.org/10.56578/jimd030304
- Views 49
- Downloads 0
A Fuzzy Similarity Based Classification with Archimedean-Dombi Aggregation Operator
The term "classification" refers to a supervised learning technique in which samples are given class labels based on predetermined classes. Fuzzy classifiers are renowned for their ability to address the issue of outlier...
Relationship Between International Trade and Logistics: An Evaluation on Countries of Shanghai Pact and the Belt and Road Initiative
Although some of the resources required to meet human needs can be provided by a country, the rest must be supplied from other countries because not every country has all resources. Therefore, the demand for internationa...
High-Performance Carbon Cycle Supply Data Sharing Method Based on Blockchain Multichain Technology
In the evolution of blockchain technology, the traditional single-chain structure has faced significant challenges, including low throughput, high latency, and limited scalability. This paper focuses on leveraging multic...
Evaluating Governance Models in Intermodal Terminal Operations: A Hybrid Grey MCDM Approach
Intermodal transportation, crucial for contemporary logistics, enhances supply chain efficiency through integrated multimodal coordination. Central to this ecosystem, intermodal terminals act as pivotal points for seamle...
Evaluating Supply Chain Efficiency Under Uncertainty: An Integration of Rough Set Theory and Data Envelopment Analysis
The evaluation of supply chain (SC) efficiency in the presence of uncertainty presents significant challenges due to the multi-criteria nature of SC performance and the inherent ambiguities in both input and output data....