A retail company needs its public service and its release pipeline to survive a major incident in one Azure region. Which architecture meets this goal?
Show answer and explanation
Answer: C
Running the service in several regions with health-checked global routing lets Azure Front Door or Traffic Manager stop sending traffic to a failed region and shift it to healthy ones. Active-active or active-passive designs both qualify, and the release process can target the surviving regions. A single region fails with its outage, pinned DNS cannot redirect clients, and a backup restore takes far longer than automatic failover.
Why the other options are wrong
- A. A larger VM in one region adds capacity but still shares the same region's failure domain, so a regional outage takes everything down together.
- B. Hard-coded records pointing at one regional load balancer send every client to a dead endpoint during an outage, because nothing re-routes them.
- D. Agents and targets in a single region keep the outage's blast radius, and manual restores take hours, which misses the continuity requirement.