EXPLORE PUBLICATIONS BY COUNTRIES

Microservices, Monolithic Architecture, Migration Styles, Strangler Fig Pattern, Refactoring Software.

The world-wide migration to microservices from monolithic is driven by demands for scalability, agility, and maintainability.1 Migration as such is however afflicted with technical and organizational issues.2 The choice of the optimal migration approach will be the recipe for success, but little empirical data to compare it with exists. The research proposes a comparative investigation of the three most debated migration approaches: Big Bang, Strangler Fig, and Parallel Run. We are comparing their impact on some of the most critical performance metrics like downtime, migration expense, project time, and system resilience. The research is done using simulated data of 471 small to large enterprise migration projects. Python Pandas libraries were utilized in data handling and Scikit-learn for use of the regression model for strategy selection and project performance forecasting. Deployment trends were emulated by Docker and Kubernetes for ensuring resiliency capability. Following our findings, in some cases Big Bang style is quicker, the Strangler Fig approach is always lower in operational risk and higher in resilience over the long term but at the cost of longer project duration. Parallel Run strategy is the most secure but at astronomically high cost of infrastructure. This paper suggests a quantitative method in order to enable organizations to take the correct migration strategy depending on risk tolerance and business requirements.

This study addresses the research question: "How do Big Bang, Strangler Fig, and Parallel Run migration strategies differ in terms of cost, risk, resilience, and implementation effort when applied to typical monolithic-to-microservices transformations?" The findings provide comparative, datadriven insights that can help architects and engineering leaders make informed decisions based on risk tolerance, operational continuity needs, and economic constraints.