Why AWS runs your app in a city you'll never visit twice

When you launch anything on AWS, you first pick a Region — a physical geographic area like Northern Virginia, Frankfurt, or Tokyo. There are dozens worldwide, and they're isolated from each other on purpose: data and traffic stay inside the Region you choose unless you explicitly move them. That's why picking a Region close to your users (lower latency) and compliant with local data laws is one of the first real decisions you make.

But a Region isn't a single building. Each one is divided into Availability Zones (AZs) — separate clusters of data centers, each with its own power, cooling, and networking, placed miles apart but linked by fast, private connections. If a flood or power failure knocks out one AZ, the others keep running. A Region typically has three or more AZs.

The practical rule that flows from this: to build something that survives failure, spread your servers across multiple Availability Zones. A single server in one AZ is a single point of failure; the same app running in two or three AZs stays online even when a whole data center goes dark. This design pattern — called high availability — is the reason AZs exist and why AWS keeps them physically apart yet close enough to act as one.