Why Global Resiliency?
Resiliency isn’t just about protecting against rack or datacenter failures. Entire Azure regions can experience outages due to natural disasters or network failures. To ensure business continuity, architects must design for region-level redundancy using Azure’s region pairs and global routing services.
1. Region Pairs
Definition:
Each Azure region is paired with another within the same geography (e.g., East US ↔ West US).
Key Benefits:
-
Paired regions are prioritized for recovery during outages.
-
Updates are rolled out sequentially (not at the same time).
-
Compliance: Data residency remains within geography.
Example Pairs:
-
East US ↔ West US.
-
North Europe ↔ West Europe.
-
Southeast Asia ↔ East Asia.
Exam Tip: If the requirement says “disaster recovery while ensuring data residency within same geography” → Region Pairs.
2. Traffic Manager (DNS-Based Global Routing)
Definition:
Routes users to the most appropriate endpoint using DNS resolution.
Routing Methods:
-
Priority (failover model).
-
Weighted (A/B testing, load distribution).
-
Performance (nearest endpoint by latency).
-
Geographic (compliance/regulatory).
-
Multi-Value / Subnet (advanced scenarios).
Use Cases:
-
Global SaaS applications.
-
Multi-region failover.
-
Geographic routing for compliance.
Limitations:
-
DNS-based = not real-time. Clients cache DNS.
3. Azure Front Door (Global Application Delivery)
Definition:
Provides layer 7 global routing with edge POPs for performance + availability.
Key Features:
-
Anycast routing → directs users to closest edge.
-
Application acceleration (SSL offload, caching).
-
WAF protection (similar to App Gateway).
-
Seamless failover across regions.
Use Cases:
-
Mission-critical global web apps.
-
Multi-region active-active deployment.
-
Combine with Traffic Manager for advanced compliance.
4. Best Practices for Global Resiliency
-
Deploy apps in at least 2 regions (paired).
-
Use Traffic Manager for DNS-based failover if compliance/geographic routing is needed.
-
Use Front Door for active-active, edge-based routing and acceleration.
-
Store state in globally replicated databases (Cosmos DB, SQL Auto-Failover).
-
Ensure backup/DR strategy aligns with RTO/RPO.
Example Enterprise Scenario
A global financial platform requires:
-
Must survive complete regional outage.
-
EU customers must remain within EU data centers (compliance).
-
Customers worldwide must connect to nearest region for best performance.
Correct design:
-
Deploy workloads in North + West Europe (region pair).
-
Use Traffic Manager with Geographic routing for EU compliance.
-
Use Azure Front Door for active-active load balancing worldwide.
-
Replicate DB using SQL Auto-Failover or Cosmos DB multi-region.
Confusion Buster
-
Traffic Manager vs Front Door
-
TM = DNS-based (routing only).
-
FD = edge-based (routing + acceleration + WAF).
-
-
Region Pairs vs Availability Zones
-
RP = region-level redundancy.
-
AZ = datacenter-level redundancy within one region.
-
-
Front Door vs CDN
-
FD = full application acceleration + failover.
-
CDN = static content caching only.
-
Exam Tips
-
“Which Azure feature ensures sequential updates and DR priority?” → Region Pairs.
-
“Which service routes traffic based on geographic compliance?” → Traffic Manager.
-
“Which global routing service uses edge POPs for performance?” → Azure Front Door.
-
“Which service provides real-time failover between regions?” → Front Door (not Traffic Manager).
What to Expect in the Exam
-
Direct Q: “Which service uses DNS to route users to the nearest endpoint?” → Traffic Manager.
-
Scenario Q: “Company needs active-active failover across two regions with minimal downtime.” → Azure Front Door.
-
Scenario Q: “Organization must ensure EU users’ traffic stays in EU.” → Traffic Manager with geographic routing.
-
Trick Q: “Availability Zones protect against entire region failure.” → False (that’s region pairs).