Performance Optimization in Large-Scale Database Migration A Multi-Algorithm Assessment
Tirumala Gundala
Abstract
Database migration is a critical process for organizations transitioning to modern computing systems and cloud infrastructures. This study analyzes performance factors influencing migration outcomes using data from 200 migration projects. The research examines the impact of data size (14.35–4,897.32 GB), migration complexity (1–50 steps), and team expertise on overall success rates. Correlation analysis reveals migration complexity has a strong negative relationship with success rates (-0.65), team expertise has a positive effect (0.51), and data size shows a weaker negative impact (-0.36). Random Forest Regression demonstrated the best performance on training data (R² = 0.9767, MSE = 3.7261), followed by Support Vector Regression (R² = 0.8791, MSE = 19.3033) and Linear Regression (R² = 0.8634, MSE = 21.8080). Test results were consistent, with R² values ranging from 0.7590 to 0.8077, confirming model reliability. Findings highlight team expertise as the most influential factor, enabling mitigation of challenges posed by large data volumes and complex migration processes. Migration complexity remains the primary obstacle, underscoring the need for streamlined strategies. These insights provide practical guidance for organizations, emphasizing skilled teams and simplified architectures to optimize success in cloud migration projects.