MLA-C01 Deployment and Orchestration of ML Workflows Practice Question
A company needs to deploy a new model version to a SageMaker real-time endpoint. They want to route 5% of traffic to the new version initially to monitor for errors before full rollout. Which deployment strategy should they use?
⚠ Common exam trap
Many candidates confuse canary deployment with shadow testing, mistakenly thinking shadow testing also routes live user traffic, when in fact shadow testing only duplicates traffic for validation without affecting the user experience.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Canary deployment with production variants
A canary deployment with production variants allows you to route a specific percentage of traffic (e.g., 5%) to the new model version by adjusting the `InitialVariantWeight` parameter in the production variant configuration. This enables gradual traffic shifting while monitoring errors, and you can later increase the weight to 100% for full rollout. SageMaker real-time endpoints support this natively by hosting multiple model variants behind the same endpoint.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Blue/green deployment
Why it's wrong here
Blue/green deployment shifts all traffic from the old fleet to the new one at once, so it cannot hold 5% on the new version while 95% remains on the old. It is tempting because it is a recognised SageMaker deployment option, but it would be correct when a full cutover with instant rollback is required.
- ✗
Shadow testing
Why it's wrong here
Shadow testing mirrors live traffic to the new model without returning its responses to users, so it cannot serve 5% of production requests. It is tempting because it also limits production risk, but shadow testing would be correct when validating a model's predictions against live traffic before any user exposure.
- ✓
Canary deployment with production variants
Why this is correct
Canary deployment with production variants lets SageMaker split endpoint traffic, sending 5% to the new model variant while the old version serves the rest. This directly satisfies the requirement to monitor errors before full rollout.
- ✗
Multi-model endpoint
Why it's wrong here
A multi-model endpoint hosts several models behind one endpoint but selects a model per invocation via a target model header, not by splitting traffic percentages. It is tempting because it deploys multiple models together, but it would be correct when hosting many models cost-effectively rather than gradually shifting traffic.
Go deeper
Related to this question
About these practice questions
One of 665 original MLA-C01 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This MLA-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the MLA-C01 exam.