Availability Zone
Definition
A physically separate data center within a cloud region with independent power and cooling, connected to peer AZs for fault-isolated redundancy.
Use Cases
- Netflix: Keep video streaming available during data center failures and maintenance events. — Runs production workloads on AWS across multiple Availability Zones within a region, using load balancing and automated scaling so traffic can shift away from an impaired zone. (Improved resilience and reduced customer impact from infrastructure failures by avoiding single–data center dependencies.)
- Airbnb: Maintain uptime for booking and search services during infrastructure disruptions. — Uses AWS multi-AZ patterns for critical services (for example, deploying application tiers across multiple Availability Zones and using managed databases with Multi-AZ capabilities). (Higher service availability and better fault tolerance for customer-facing systems.)
- Spotify: Ensure continuous music playback and platform availability during outages and routine maintenance. — Operates on Google Cloud with regional architectures that distribute services across multiple zones, combined with load balancing and automated recovery. (Reduced risk of downtime from single-zone failures and smoother maintenance with less user disruption.)
Provider Equivalents
- AWS: Availability Zone
- Azure: Availability Zone
- GCP: Zone
- OCI: Availability Domain
Frequently Asked Questions
- What's the difference between an Availability Zone and a Region?
- A Region is a larger geographic area (like a metro area or country/area) that contains multiple isolated locations. An Availability Zone (AZ) is one of those isolated locations—typically one or more data centers with independent power, cooling, and networking. You use multiple AZs inside one Region to survive a data-center-level failure without moving to a different geography.
- When should I use multiple Availability Zones?
- Use multiple AZs when the application must stay online during a data center outage or when you need to perform maintenance with minimal downtime. Common cases include production web apps, APIs, and databases that require high availability. For dev/test or non-critical batch jobs, a single AZ may be acceptable if downtime is tolerable.
- How much does using multiple Availability Zones cost?
- There is no separate fee just to 'use an AZ,' but costs often increase because you run duplicate resources (for example, instances in two AZs) and may pay for cross-AZ data transfer. Pricing depends on the provider and service: compute and database replicas add hourly charges, and network traffic between AZs can add per-GB costs. The main cost drivers are redundancy (more resources) and cross-zone traffic volume.
Category: cloud
Difficulty: intermediate
Related Terms
See Also