This section describes an availability zone and region as used for stretch clusters.

Availability Zones

An availability zone is a collection of infrastructure components. Each availability zone runs on its own physically distinct, independent infrastructure, and is engineered to be highly reliable. Each zone should have independent power, cooling, network, and security.

Additionally, these zones should be physically separate so that disasters affect only one zone. The physical distance between availability zones is short enough to offer low, single-digit latency (less than 5 ms) and large bandwidth (10 Gbps) between the zones.

Availability zones can either be two distinct data centers in a metro distance, or two safety or fire sectors (data halls) in the same large-scale data center.

Regions

Regions are in two distinct locations - for example, region A can be in San Francisco and region B in Los Angeles (LAX). The distance between regions can be rather large. The latency between regions must be less than 150 ms.