MLA-C01 Deployment and orchestration of ML workflows

A team wants to send a small first slice of production traffic to a retrained recommendation model, watch error-rate and latency alarms during a bake period, and then shift all remaining traffic at once if everything looks fine. Which SageMaker traffic shifting mode is described?

Show answer and explanation

Answer: D

Canary mode in a SageMaker blue/green deployment shifts a small first portion of traffic to the new fleet, holds it for the bake period while CloudWatch alarms are monitored, and then shifts the remaining traffic in a single step. Alarms can trigger rollback. Linear mode instead moves traffic in several equal increments, and an all-at-once switch has no staged exposure.

Why the other options are wrong

  • A. An all-at-once switch sends every request to the new model simultaneously. The scenario requires limiting exposure first, so an immediate full cutover fails the staged requirement.
  • B. Linear mode moves traffic in several equal-sized steps. The scenario describes one small first slice followed by a single shift of the rest, which is the canary pattern instead.
  • C. Offline batch scoring computes predictions for stored datasets rather than live requests. It cannot test the new model against real user traffic with a staged live exposure.

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.