An Approach for Automated Vulnerability Analysis of Apache Spark-Based Computational Clusters in Kubernetes Environments

This paper presents a software module designed to enhance the security analysis of Apache Spark-based computational clusters deployed in Kubernetes orchestration environments. Current automated vulnerability detection tools for Kubernetes and Apache Spark are analyzed, with an emphasis on their features, strengths, and limitations. A comprehensive security assessment was performed using vulnerability scanning and penetration testing tools, such as Nmap. Based on the evaluation results, an integrated software framework was developed that consolidates outputs from multiple analyzers into a unified report. Additionally, a mathematical model for evaluating the effectiveness of security analysis is proposed, incorporating metrics on the number and severity of detected vulnerabilities, as well as the time required for scanning. This model facilitates a quantitative comparison of tool performance and highlights the advantages of the proposed module. The architectural design and implementation details of the module are provided, along with recommendations for mitigating identified vulnerabilities. Experimental results demonstrate improved efficiency and accuracy in cluster security analysis.