An availability zone is a physically separate data center or group of data centers within a cloud region that is designed to fail independently from others. Each zone has its own power, cooling, and networking, while still being connected to other zones with high speed links. Cloud providers encourage you to deploy virtual machines, databases, and other resources across multiple zones to improve reliability. When you choose a region, you usually see several availability zones labeled with suffixes like a, b, or c.
Why it matters
Designing for high availability means assuming that any single server or even an entire building can fail. By spreading workloads across more than one availability zone, you can survive many local outages without downtime. This is especially important for critical web applications where users expect near constant uptime.
How it works
Load balanced traffic is directed to healthy instances in multiple zones, and data is often replicated across zones using managed relational databases or object storage. Some services automatically handle cross zone replication, while others require explicit configuration. You can explore the bigger picture in the lessons Regions and Availability Zones and High Availability.