If you are new to cloud infrastructure, the geography vocabulary is one of the first places the marketing and the engineering diverge. It is worth getting straight, because resilience and data residency both depend on it.
Region
A region is a geographic area containing multiple isolated data centres — Mumbai, Doha, Frankfurt. It is the unit of data residency: resources in a region stay in that region unless you deliberately move them.
It is also the unit of pricing and of service availability. Not every service exists in every region, which catches out architectures designed against documentation written for a different one.
Availability zone
An availability zone is one or more physically separate data centres inside a region, with independent power, cooling, and networking. Zones within a region are close enough for fast replication and far enough apart that one flooding does not take out the others.
Running across two zones is the default answer to "how do I survive a data centre failure". It is usually cheap and frequently skipped.
Edge location
Edge locations cache content close to users. They serve traffic; they are not where your data lives, and they are not a resilience mechanism for your application.
Why the distinction matters
| Concern | Handled by |
|---|---|
| A server dies | Multiple instances |
| A data centre fails | Multiple availability zones |
| A whole region fails | Multi-region — expensive, rarely justified |
| Users are far away | Edge locations / CDN |
| Data must stay in-country | Region selection |
Most organisations need multi-zone and do not need multi-region. Being clear about which problem you are solving keeps you from paying for the wrong one.