Akıllı sistemler ve uygulamaları dergisi

Confidential state databases are increasingly being migrated to managed cloud platforms such as AWS RDS for better scalability, resilience, and administrative control. However, this transition also introduces security risk because encryption states, access paths, and temporary exposure conditions can change during migration. Existing studies discuss cloud security and access control, but they do not clearly explain how confidentiality risk develops across source, transit, staging, and target layers during migration of highly sensitive databases. This study addresses that gap by treating AWS RDS migration as a controlled security process and evaluating encryption boundary integrity, access control deviations, privilege propagation, and containment response under simulated failure conditions. The results show that the highest risk appears in intermediate migration stages, especially staging, while the final controlled AWS RDS state provides stronger confidentiality stability than the legacy environment. The study concludes that secure migration depends on continuous encryption enforcement, strict privilege segmentation, and rapid containment of transitional exposure. These findings are useful for government and other high-sensitivity sectors that need secure database modernization.