# Availability Zone
An Availability Zone (AZ) is a distinct, physically isolated location within a cloud provider's region, equipped with its own power, cooling, and networking. Multiple AZs in a region are interconnected with low-latency links, allowing cloud customers to build highly available and fault-tolerant applications that can withstand data center-level failures.
**Pronunciation:** uh-vey-luh-bil-i-tee zohn
**Difficulty:** Beginner
**Synonyms:** AZ, Fault Domain
**Categories:** Cloud Infrastructure, System Design
**Tags:** cloud computing, infrastructure, high availability, fault tolerance, disaster recovery, aws, azure, gcp
Canonical: https://scaleengineer.com/glossaries/availability-zone
---
## Definitions

- **Core Concept in Cloud Infrastructure:** An **Availability Zone (AZ)** is a fundamental building block of cloud infrastructure, representing one or more discrete data centers with redundant power, networking, and cooling, all housed in separate facilities within a specific geographic region.

### Key Concepts
*   **Isolation**: **AZs** are physically separate, meaning a failure in one **AZ** (e.g., fire, power outage, flood) is highly unlikely to affect others in the same region. This isolation is the cornerstone of building fault-tolerant systems. A similar concept in some platforms is a **Fault Domain**.
*   **Low-Latency Connectivity**: All **AZs** within a single cloud region are interconnected with high-bandwidth, low-latency private networking. This allows for fast, synchronous data replication and communication between zones, which is crucial for stateful applications like databases that require rapid failover.
*   **Redundancy and High Availability**: The core principle is to run applications and store data across multiple **AZs**. If one **AZ** becomes unavailable, traffic can be automatically rerouted to the healthy **AZs**, ensuring continuous operation and minimizing downtime.

### Example Usage
A common high-availability architecture involves deploying a web application on virtual machines in at least two **AZs**. An Elastic Load Balancer distributes incoming traffic between them. If the instances in one **AZ** fail, the load balancer automatically directs all traffic to the instances in the other **AZ**, maintaining service availability for users.

Similarly, a managed database service might be configured with a primary instance in one **AZ** and a synchronous standby replica in another. If the primary instance fails, the service automatically promotes the standby to become the new primary, typically with minimal to no data loss.

## Etymology

The term combines 'Availability,' referring to system uptime and accessibility, with 'Zone,' denoting a distinct physical or logical area. It was coined by cloud providers to describe a specific, isolated infrastructure unit designed to ensure the continuous availability of services.

## First used

2006

## Historical context

The concept of the **Availability Zone** emerged with the rise of modern cloud computing, pioneered by Amazon Web Services (AWS) around 2006. Before the cloud era, achieving high availability required businesses to build and manage multiple, geographically dispersed data centers—a complex and expensive endeavor.

AWS introduced the model of Regions and **Availability Zones** to democratize this capability. A Region represents a broad geographic area (e.g., us-east-1), while an **AZ** is an isolated, independent data center or cluster of data centers within that Region. This architecture was designed to solve the inherent problem of single-data-center deployments being vulnerable to localized failures.

By offering multiple **AZs** connected by low-latency networks, AWS enabled developers to easily build fault-tolerant applications that could survive the failure of an entire data center. This model proved so effective that other major cloud providers, including Google Cloud Platform (GCP) and Microsoft Azure, adopted similar constructs. Azure, for instance, uses the terms **Availability Zones** and **Fault Domains** to describe similar isolation concepts. This architectural pattern is now a fundamental standard for designing resilient and highly available systems in the cloud.

## Q&A

- **What is the primary purpose of using multiple Availability Zones in a cloud architecture?:** The primary purpose is to achieve high availability and fault tolerance. By distributing application components and data across multiple, isolated **Availability Zones**, an application can remain operational even if one entire **AZ** fails due to events like power outages, network failures, or natural disasters. This design eliminates the single point of failure associated with a single data center.
- **How do Availability Zones within the same region communicate with each other?:** Availability Zones within the same region are interconnected with high-bandwidth, low-latency private network links. This allows for fast and efficient communication between resources in different zones, enabling critical operations like synchronous data replication for databases and rapid failover without significant performance degradation.

## Usage examples

- To ensure our application can withstand a data center outage, we deployed our database replicas across three different **Availability Zones** within the same region.
- The system architect decided to use a load balancer to distribute traffic evenly across web servers located in separate **AZs**, enhancing the system's resilience.
- In Microsoft Azure, a **Fault Domain** serves a similar purpose to an **Availability Zone**, providing physical isolation for virtual machines within a data center to protect against localized hardware failures.

## Related terms

- Region
- High Availability
- Fault Tolerance
- Disaster Recovery
- Load Balancing
- Data Center

## Popular related terms

- Region
- High Availability
- Fault Tolerance