Overview
Microsoft Azure is one of the largest cloud platforms in the world, with a massive global physical infrastructure designed for reliability, performance, and data residency compliance.
Understanding how Azure organizes its global infrastructure is essential — both for the AZ-900 exam and for real-world Azure deployments.
The hierarchy looks like this:
Geographies
└── Regions
└── Availability Zones
└── Data Centers
Geographies
A geography is a defined area of the world that contains one or more Azure regions. Geographies are designed to respect data residency, sovereignty, and compliance boundaries.
Examples of Azure Geographies:
-
Americas
-
Europe
-
Asia Pacific
-
Middle East and Africa
-
Australia
-
Brazil
-
India
Each geography has at least two regions (for redundancy), and data generally stays within the geography unless explicitly configured otherwise.
Why it matters: Some countries require that data about their citizens never leaves national borders. Azure geographies allow compliance with laws like GDPR (Europe) and India's PDPB.
Regions
An Azure region is a geographic area containing one or more data centers that are networked together with a low-latency network.
Key Facts About Regions
|
Fact |
Detail |
|---|---|
|
Total regions |
60+ worldwide (more than any other cloud provider) |
|
How you use them |
You choose a region when deploying resources |
|
Latency |
Resources in the same region communicate with low latency |
|
Pricing |
Prices vary slightly between regions |
|
Data residency |
Your data stays in your chosen region by default |
Examples of Azure Regions
|
Region Name |
Location |
|---|---|
|
East US |
Virginia, USA |
|
West Europe |
Netherlands |
|
Southeast Asia |
Singapore |
|
Central India |
Pune, India |
|
Australia East |
New South Wales, Australia |
|
UK South |
London, UK |
|
Brazil South |
São Paulo, Brazil |
How to Choose a Region
When deploying resources, consider:
-
Proximity to users — deploy close to where your users are for low latency
-
Available services — not all services are available in every region
-
Compliance — data residency laws may restrict your region choice
-
Pricing — costs can vary between regions
Availability Zones
Availability Zones (AZs) are physically separate data centers within a single Azure region — each with its own power, cooling, and networking.
Azure Region: East US
├── Availability Zone 1 (Data Center A)
│ Power: Independent
│ Cooling: Independent
│ Network: Independent
├── Availability Zone 2 (Data Center B)
│ Power: Independent
│ Cooling: Independent
│ Network: Independent
└── Availability Zone 3 (Data Center C)
Power: Independent
Cooling: Independent
Network: Independent
Why Availability Zones Exist
If one data center has a power outage, a cooling failure, or any physical disaster — the other zones continue running normally. Your application stays up.
How AZs Protect You
By deploying your application across multiple zones, you achieve:
-
Zero downtime during single-zone failures
-
99.99% SLA for VM deployments across zones
-
Protection from local disasters (flood, fire, power failure in one building)
Zone-Enabled Azure Services Examples
|
Service |
Zone Support |
|---|---|
|
Azure Virtual Machines |
Zone-specific deployment |
|
Azure Load Balancer (Standard) |
Zone-redundant |
|
Azure SQL Database |
Zone-redundant option |
|
Azure Kubernetes Service |
Zone-aware node pools |
|
Azure Storage |
Zone-redundant storage (ZRS) |
AZ vs. No AZ
Single Zone Deployment:
┌──────────┐
│ Zone 1 │ ← If this fails, app goes down
└──────────┘
Multi-Zone Deployment:
┌──────────┐ ┌──────────┐ ┌──────────┐
│ Zone 1 │ │ Zone 2 │ │ Zone 3 │
└──────────┘ └──────────┘ └──────────┘
↑ If Zone 1 fails, Zone 2 and Zone 3 keep serving traffic ↑
Region Pairs
Every Azure region is paired with another region within the same geography — at least 300 miles (480 km) apart. This is called a region pair.
Purpose of Region Pairs
|
Benefit |
Detail |
|---|---|
|
Disaster recovery |
If an entire region fails, the paired region takes over |
|
Sequential updates |
Azure updates paired regions one at a time to prevent simultaneous outages |
|
Data residency |
Data replication stays within the same geography |
|
Priority recovery |
If a major disaster affects multiple regions, one region in each pair is prioritized for recovery |
Example Region Pairs
|
Primary Region |
Paired Region |
|---|---|
|
East US |
West US |
|
North Europe (Ireland) |
West Europe (Netherlands) |
|
Southeast Asia (Singapore) |
East Asia (Hong Kong) |
|
Central India |
South India |
|
UK South |
UK West |
|
Australia East |
Australia Southeast |
Region Pair vs. Availability Zone
|
|
Availability Zone |
Region Pair |
|---|---|---|
|
Scope |
Within one region |
Across two regions |
|
Distance |
A few miles apart |
300+ miles apart |
|
Protects against |
Data center failure |
Regional disaster |
|
Latency |
Very low |
Higher |
|
SLA |
99.99% |
Used for DR strategy |
Sovereign Regions
Azure also operates sovereign cloud regions — physically and logically isolated from the public Azure cloud — for government and highly regulated customers.
|
Sovereign Cloud |
Who It's For |
|---|---|
|
Azure Government |
US federal, state, and local government agencies |
|
Azure China |
Chinese organizations (operated by 21Vianet, a local partner) |
|
Azure Germany |
German government and data residency requirements |
Sovereign clouds have separate portals, APIs, and physical infrastructure. Not all Azure services are available in all sovereign clouds.
Infrastructure Summary
Azure Global Infrastructure:
─────────────────────────────
Geographies → 30+ (Americas, Europe, Asia, etc.)
Regions → 60+ globally
Availability Zones → 3 per supported region
Region Pairs → Every region paired with another
Why it matters:
✓ Deploy near your users (low latency)
✓ Meet data residency laws (geography/region)
✓ Survive data center failures (Availability Zones)
✓ Survive regional disasters (Region Pairs)
Official References
Next Chapter → Chapter 08: Azure Resources, Resource Groups, Subscriptions & Management Groups