Task 3.2: Create and Script Infrastructure Based on Existing Architecture and Requirements MLA-C01 Practice Questions (Page 6)

Questions 26–30

1. 26foundation · easy

   Which AWS service is a fully managed container registry for storing and retrieving container images?

   • AAWS Fargate
   • BAmazon ECS
   • CAmazon ECR
   • DAmazon EKS
   Reveal Solution Discussion Select an answer first
2. 27expert · hard

   A company runs a SageMaker AI endpoint for a model that has a predictable daily pattern with a peak at noon. They also experience occasional, unpredictable spikes. They want to minimize latency during both the predictable peak and the unpredictable spikes while keeping costs low. Which scaling policy combination should they use?

   • AOnly scheduled scaling for the noon peak
   • BManual scaling with a fixed number of instances
   • CScheduled scaling for the noon peak and target tracking for the unpredictable spikes
   • DOnly target tracking for all traffic
   Reveal Solution Discussion Select an answer first
4. 28foundation · easy

   What is the role of a container orchestrator like Amazon ECS or Amazon EKS?

   • ATo provide a virtual machine for each container
   • BTo store container images in a central repository
   • CTo build container images from a Dockerfile
   • DTo manage the deployment, scaling, and operation of containers across a cluster
   Reveal Solution Discussion Select an answer first
5. 29foundation · easy

   What is a primary benefit of using AWS CloudFormation for ML infrastructure?

   • AIt automatically optimizes the cost of all deployed resources.
   • BIt provides a visual drag-and-drop interface for designing infrastructure.
   • CIt eliminates the need to write any configuration for resources.
   • DIt allows you to manage infrastructure as code, enabling version control and repeatable deployments.
   Reveal Solution Discussion Select an answer first
6. 30application · medium

   An ML team wants to use a custom container for SageMaker AI training. They have built the Docker image locally and tested it. What is the next step to make this image available for SageMaker training jobs?

   • ACopy the image to an EC2 instance and run the training job there
   • BPush the image to Amazon ECR and reference the ECR image URI in the training job
   • CUpload the Docker image to S3 and specify the S3 path in the training job
   • DRegister the image in AWS Marketplace and use it in SageMaker
   Reveal Solution Discussion Select an answer first