Simulation-Driven Optimization of Containerized Systems Using the ContSysSim Framework

Containerized microservice architectures have become prevalent in modern distributed systems, yet accurate resource sizing remains challenging due to fluctuating workloads. This paper presents ContSysSim, an open-source Python framework that enables simulation-driven optimization of containerized systems through discrete-event modeling with integrated fluctuation effects and temporal resource reservations. We introduce a workflow combining real-world measurement collection via an Analyzer component with simulation to support infrastructure design decisions. Experimental validation demonstrates 26% resource reduction while maintaining performance. Case studies show that ContSysSim effectively models resource patterns and enables data-driven capacity planning for containerized infrastructures.