MLA-C01 Deployment and orchestration of ML workflows

A trading platform needs model inference to survive the loss of an entire AWS Region, with automatic failover and no manual runbook. Which architecture should be used?

Show answer and explanation

Answer: B

Regional resilience requires the model artifacts, endpoint configuration and supporting data to exist in a second region, because a SageMaker endpoint cannot span regions. Route 53 health checks then remove the failed region from DNS and send traffic to the survivor. Active-active serves from both regions and needs data and model parity, while active-passive keeps a warm standby that is scaled up on failover. Recovery objectives determine which variant is appropriate, and both must be tested.

Why the other options are wrong

  • A. A large single-region deployment gives no protection against a regional outage, which is the scenario being planned for.
  • C. Manual switching depends on a person noticing and acting, so it cannot meet a no-runbook requirement.
  • D. Batch transform produces scheduled offline predictions and cannot take over live request serving.

Keep going. 212 MLA-C01 questions, answers explained after each one.

Practise this domain

Reviewed November 2026. Spotted a problem with this question? Report it from the practice tool.