Availability Zone

An availability zone is a physically isolated location within a cloud provider's region, designed to operate independently from other zones in the same region. Each zone has its own power, cooling, and networking infrastructure, so that a failure in one zone does not cascade to others. The availability zone is the fundamental unit of Fault tolerance in cloud architecture: it is not a failure prevention mechanism but a failure containment boundary.

The design of availability zones reflects a specific theory of distributed systems: that correlated failures are more dangerous than independent failures, and that the best way to survive correlated failures is to isolate them physically. This is why cloud providers locate availability zones kilometers apart — not merely for redundancy, but for correlated failure independence. The cost is latency: synchronizing data across zones introduces network delays that constrain the consistency guarantees the system can offer.