SkarpChapter 6 of 20
Azure Global Infrastructure: Regions, Region Pairs, and Availability Zones
Follow your workloads across the globe to see how Azure regions and Availability Zones work together to deliver resiliency, performance, and data residency.
Big Picture: Why Azure Global Infrastructure Matters
From Cloud Theory to Geography
You know cloud basics and the shared responsibility model. Now we focus on Azure's physical and geographic layer: how Microsoft arranges datacenters worldwide.
Hidden Requirements in Scenarios
Exam questions often hide infrastructure ideas inside business needs: low latency, data residency, or staying online during failures or maintenance.
Three Core Concepts
You must understand: Azure regions (where you deploy), region pairs (how Azure links regions), and Availability Zones (separate locations within a region).
Your Goal in This Module
By the end, you should be able to choose regions and zones that best match availability, compliance, and performance requirements in AZ-900 scenarios.
Azure Regions: The Basic Unit of Geography
What Is an Azure Region?
An Azure region is a set of datacenters in a specific geographic area, connected by a low-latency network. It is the named location where you deploy Azure resources.
Region Examples
Examples include East US, West Europe, North Europe, Japan East, and Brazil South. Each is a distinct deployment location you choose at resource creation.
Why Region Choice Matters
Region affects three big things: which services are available, how fast users reach them (latency), and where data is stored for residency and compliance.
Regions vs Geographies
Microsoft also defines geographies (like Europe or Asia Pacific) that contain multiple regions and support data residency commitments inside that geopolitical area.
Choosing a Region: Latency and Data Residency Scenarios
Scenario 1: EU Startup
EU startup with EU customers needs low latency and EU data residency. West Europe or North Europe are strong choices inside the Europe geography.
Scenario 2: US-Focused App
Global app but 80% of users in eastern US. Choosing East US or East US 2 places compute near most users, reducing latency.
Scenario 3: Public Sector
Some government customers must use specialized regions (such as US Gov). Here, compliance and residency can outweigh pure performance.
Exam Pattern
On AZ-900, when you read 'must keep data in X' or 'low latency for users in Y', you are being tested on picking an appropriate Azure region or geography.
Region Pairs: Built-In Disaster Recovery Design
What Is a Region Pair?
A region pair is two Azure regions in the same geography, linked for business continuity and disaster recovery. They are directly connected and designed to work together.
Key Property: Same Geography
Both regions in a pair are in the same geography, helping meet data residency and compliance needs while still providing cross-region redundancy.
Separated but Connected
Paired regions are separated by hundreds of kilometers, reducing the chance that the same natural or regional disaster impacts both at once.
Update Sequencing and Recovery
Azure rolls out platform updates to one region in a pair at a time and prioritizes at least one region in each pair for recovery during broad outages.
Region Pairs in Action: Disaster Recovery Designs
Single Region vs Paired Region
If you deploy only in West Europe, a regional outage takes your app down. Using its region pair (North Europe) gives you a backup location.
Design B: Paired DR Setup
Primary in West Europe, replicas and backups in North Europe. Use Traffic Manager or Front Door to fail over if the primary region fails.
Benefits of Using the Pair
You gain resiliency, stay aligned with Azure's geo-redundant features, and keep data inside the same geography for residency.
Exam Trap to Avoid
If a question mentions disaster recovery plus EU data residency, do not pick a non-EU region. The secondary should be the paired region in the same geography.
Availability Zones: High Availability Inside a Region
What Is an Availability Zone?
An Availability Zone is a unique physical location within a region, made up of one or more datacenters with independent power, cooling, and networking.
Physical Separation
Zones are designed so that a failure in one zone, such as a power issue, should not take down the other zones in the same region.
Zone-Aware Services
Some services are deployed in a specific zone, others are zone-redundant and automatically spread across zones for higher availability.
Zones vs Regions
Zones protect against datacenter-level failures inside a region. To handle full regional outages, you still need cross-region replication using region pairs.
Designing with Availability Zones: Three-Tier App
Three-Tier App in One Zone
If all tiers of your app end up in the same datacenter, a single building-level failure can take the entire application offline.
Spreading Across Zones
Place VMs and services across zone 1 and zone 2. If one zone fails, the other can keep serving traffic with help from load balancing.
Impact on SLAs
Many Azure services offer higher SLAs when you use zone-redundant configurations, because they can survive zone-level failures.
Sets vs Zones
Availability Sets are within one datacenter; Availability Zones are separate datacenters. Zones provide stronger isolation and resilience.
Thought Exercise: Match Requirements to Regions and Zones
Work through these mini-scenarios. Do not worry about exact region names; focus on reasoning. After you think through each, compare with the guiding answers.
Scenario A: Banking app in Canada
- Requirements:
- Data must remain in Canada for regulatory reasons.
- Very high availability within the country.
- Your design:
- Which geography/region(s) would you consider?
- Would you use Availability Zones, region pairs, or both?
Scenario B: Internal HR portal for a small company in Singapore
- Requirements:
- Staff are mainly in Singapore.
- Outage of a few hours is acceptable.
- Budget is tight; they want to keep architecture simple.
- Your design:
- Would you prioritize multiple regions or a single region close to users?
- Would you pay extra for cross-region replication?
Scenario C: Global SaaS with strict uptime
- Requirements:
- Users in North America and Europe.
- Downtime must be minimized, even during regional disasters.
- Your design:
- How would you place deployments across regions and region pairs?
- How would you use Availability Zones within each region?
Guiding answers (self-check):
- A: Use Canadian regions (for example, Canada Central and its pair), with zone-redundant design and cross-region DR.
- B: Likely a single region near Singapore, possibly without cross-region DR to save cost; zones optional depending on SLA.
- C: Multi-region deployments in paired regions in both North America and Europe, each region using Availability Zones for intra-region resilience.
Quick Check 1: Regions and Region Pairs
Test your understanding of Azure regions and region pairs.
A company runs its main workload in West Europe and must implement cross-region disaster recovery while keeping data within the same geography. Which region is the MOST appropriate target for replication?
- East US
- North Europe
- Australia East
- Central India
Show Answer
Answer: B) North Europe
North Europe is the correct-style answer because it is in the same Europe geography and is part of the region pair with West Europe. East US, Australia East, and Central India are in different geographies and would not satisfy the 'same geography' data residency requirement.
Quick Check 2: Availability Zones vs Regions
Test your understanding of Availability Zones.
What is the PRIMARY purpose of using Availability Zones within a single Azure region?
- To ensure data residency across multiple countries
- To protect workloads against datacenter-level failures
- To provide lower latency to users in different continents
- To reduce the cost of running virtual machines
Show Answer
Answer: B) To protect workloads against datacenter-level failures
Availability Zones are physically separate locations within a region designed to protect against datacenter-level failures. They do not provide cross-country residency (that is about regions/geographies) or global latency improvements (that is about choosing regions close to users). Costs may increase rather than decrease when using zones.
Key Term Review: Regions, Region Pairs, Availability Zones
Flip these cards (mentally) to reinforce the core concepts before moving on.
- Azure region
- A set of datacenters deployed within a specific geographic area and connected through a dedicated, low-latency network; the named location where you deploy Azure resources.
- Region pair
- Two Azure regions within the same geography that are directly connected and designed to support business continuity and disaster recovery, with separated locations and sequenced platform updates.
- Availability Zone
- A unique physical location within an Azure region, made up of one or more datacenters with independent power, cooling, and networking, used to increase availability and fault tolerance inside the region.
- Latency (in Azure context)
- The time it takes for data to travel between a user and an Azure service; typically reduced by placing resources in a region close to the majority of users.
- Data residency
- The geographic location where data is stored and processed, often driven by regulatory or organizational requirements; influenced by Azure region and geography selection.
- Zone-redundant service
- An Azure service configuration that automatically spreads resources across multiple Availability Zones in a region to improve availability and resilience.
- Region-level outage vs zone-level failure
- A zone-level failure impacts one Availability Zone (datacenter group) within a region; a region-level outage affects the entire region. Zones protect against the former, region pairs help with the latter.
Putting It All Together: Strategy for AZ-900 Scenarios
Step 1: Data Residency
First, decide where data is allowed to live. Pick an Azure geography and region that meet legal or organizational residency requirements.
Step 2: Latency and Users
Next, place your primary region close to the majority of users to minimize latency and improve responsiveness.
Step 3: Region Pairs for DR
If the workload is critical, add cross-region disaster recovery using the paired region in the same geography.
Step 4: Zones for High Availability
Inside your chosen region, use multiple Availability Zones when you need to survive datacenter-level failures and reach higher SLAs.
Key Terms
- Latency
- The time it takes for data to travel between a user and an Azure service; typically reduced by placing resources in a region close to the majority of users.
- Region pair
- Two Azure regions within the same geography that are directly connected and designed to support business continuity and disaster recovery, with separated locations and sequenced platform updates.
- Azure region
- A set of datacenters deployed within a specific geographic area and connected through a dedicated, low-latency network; the named location where you deploy Azure resources.
- Data residency
- The geographic location where data is stored and processed, often driven by regulatory or organizational requirements; influenced by Azure region and geography selection.
- Availability Zone
- A unique physical location within an Azure region, made up of one or more datacenters with independent power, cooling, and networking, used to increase availability and fault tolerance inside the region.
- Zone-redundant service
- An Azure service configuration that automatically spreads resources across multiple Availability Zones in a region to improve availability and resilience.
- Business continuity and disaster recovery (BCDR)
- A planning and design approach that ensures critical applications and data remain available or can be quickly restored after failures, disasters, or major outages.