Picking the right database for a cloud application sounds simple until you actually have to do it. SQL systems like MySQL have been the go-to for decades, and for good reason — but as applications started dealing with more varied data at larger scale, NoSQL systems like MongoDB began showing up in more and more production stacks. The question was never really which one is better. It’s which one actually fits your situation. This paper grew out of that frustration. We wanted something more concrete than the usual “it depends” answer, so we put both systems through real tests on the same dataset — 100,000 e-commerce records — and measured what actually happened: query speeds, read and write throughput, storage consumption, and how each system held up as data volume grew. MySQL was faster when queries got complicated, particularly joins and aggregations. MongoDB pulled ahead on write speed and handled document-style data more naturally, though it used noticeably more storage for the same records. We also cover the theory — ACID, CAP, BASE — not to pad the paper, but because those concepts genuinely explain why the benchmarks turned out the way they did. The goal is simple: give developers and architects a clear enough picture to make this call confidently for their own application.