Methodology for Quantitative Assessment of Quality Characteristics in Distributed Systems and Microservices

The thesis discusses a new methodological approach called flexible quality assessment, aimed at improving the methodology based on the adaptive quality assessment approach used to address issues related to the uncertainty of quantitative assessment of quality characteristics in distributed systems and microservices. It analyzes the shortcomings of the adaptive quality assessment method and presents a mathematical framework for the new methodological solution, in which static probabilities are replaced by Bayesian networks that update probabilities based on new data. The mathematical model is developed based on the combined application of classical aspects of probability theory, Bayesian networks, the use of a semi-iterative approach for updating weights and assessments based on new data, as well as the use of DevOps practices for automating the deployment, monitoring, and updating of the system. The advantage of this approach is that the continuous refinement of the model based on new data allows for more accurate consideration of the dependencies between system components and their impact on reliability.