Misconfiguration of Cluster and IoT Systems Recovery: Extended Experiments

Abstract

Containerized cluster systems and edge devices are vulnerable to security breaches when configuration errors occur. In this paper, we propose more experiments on the self-healing approach for misconfigurations that adopt a Markov Decision Process to determine the optimal recovery action or policy that maximizes performance metrics. The experiments conducted are an improvement on our previous self-healing approach. Our findings demonstrate the effectiveness of the proposed controller in enhancing performance and reliability, leading to precise and improved results.