Why Disaster Recovery (DR)?
Backups protect against data loss, but what about service continuity when an entire region goes down? That’s where Disaster Recovery (DR) comes in. Azure Site Recovery (ASR) ensures that workloads can failover to another region or site quickly with minimal downtime.
Azure Site Recovery (ASR) Overview
Definition:
-
A native Azure service that replicates VMs and workloads from one site (on-prem or Azure) to another Azure region.
Key Features:
-
Continuous replication of VMs (Azure, Hyper-V, VMware, on-prem).
-
Application-consistent snapshots (not just crash-consistent).
-
Orchestrated failover & failback.
-
Recovery plans with automation (scripts, runbooks).
-
Built-in testing without impacting production.
DR Design Considerations
-
RTO and RPO
-
RTO depends on failover automation speed.
-
RPO depends on replication frequency (as low as seconds).
-
Target Location
-
Choose paired Azure region for replication.
-
Consider compliance requirements for data residency.
-
Failover Scenarios
-
Planned Failover: Used for migrations or maintenance (no data loss).
-
Unplanned Failover: For outages/disasters (some data loss possible).
-
Cost Management
-
Replication incurs storage and network costs.
-
Compute (VMs) only billed after failover.
-
Testing
-
ASR supports non-disruptive test failovers to validate DR plan.
Example Enterprise Scenario
A global bank requires:
-
Mission-critical payment app must run even if primary Azure region fails.
-
Database must maintain RPO < 5 minutes.
-
Compliance requires workloads remain in specific geographies.
Correct design:
-
Use ASR to replicate VMs to paired Azure region.
-
Configure low RPO replication policy.
-
Define recovery plans to automate failover and failback.
-
Conduct test failovers quarterly to validate DR.
Confusion Buster
-
Backup vs ASR
-
Backup = restores data from a point in time.
-
ASR = keeps workloads continuously replicated and ready for failover.
-
-
HA vs DR
-
HA = intra-region (Availability Zones, Load Balancers).
-
DR = inter-region (ASR, replication).
-
-
Planned vs Unplanned Failover
-
Planned = no data loss, controlled.
-
Unplanned = emergency recovery, may lose recent transactions.
-
Exam Tips
-
“Which service replicates workloads to another region for DR?” → Azure Site Recovery.
-
“Which solution provides non-disruptive DR testing?” → ASR test failover.
-
“Which ensures workloads recover quickly after regional outage?” → ASR.
-
“Which protects against accidental deletion?” → Azure Backup (not ASR).
What to Expect in the Exam
-
Direct Q: “Which Azure service provides replication of workloads to another region?” → Site Recovery.
-
Scenario Q: “Company requires critical workload recovery with RPO < 5 min across regions.” → ASR.
-
Scenario Q: “Company wants to test failover without impacting production.” → ASR test failover.
-
Trick Q: “Azure Backup and Site Recovery provide the same protection.” → False.