Availability zone (AZ)
In cloud computing, an availability zone is a distinct location within a cloud provider’s infrastructure that is engineered to be isolated from failures in other availability zones.
A region is a distinct geographical area, such as a city (eg. London) or area (eg. US East). Each region contains multiple availability zones. An availability zone is analogous to a data center, which is a single physical facility. Each AZ’s data center is physically separated from the others, such that a failure in one data center is not likely to cascade into failures in other data centers within the same region.
Availability zones are used to provide high availability and fault tolerance for applications and services, even when there are data center-level failures – by doing replication across multiple availability zones within each region.
Cloud subnets are typically pinned to a single availability zone, and a cloud network spans zones within a region by joining those subnets with route tables and peering links. This keeps a failure in one zone from propagating to the others while letting traffic move between them.
Availability zones are the building block of geographic redundancy: spreading replicas and standbys across zones is what lets a service keep running through the loss of a whole data center.