AWS Certified Machine Learning – SpecialtyMachine Learning Implementation and OperationsEasy

A machine learning engineer is deploying a real-time inference endpoint for a critical fraud detection model using Amazon SageMaker. The model requires high availability and needs to be resilient to instance failures within a single Availability Zone. Which configuration should the engineer primarily focus on to address this requirement?

  1. AImplementing Elastic Inference
  2. BConfiguring a Multi-AZ endpoint
  3. CEnabling SageMaker Model Monitor
  4. DSetting up SageMaker Pipelines
Show answer & explanation

Correct answer: B. Configuring a Multi-AZ endpoint

Configuring a Multi-AZ endpoint for Amazon SageMaker ensures that the inference endpoint is deployed across multiple Availability Zones. This provides high availability and resilience against single Availability Zone outages or instance failures within a single AZ.

Why the other options are wrong

  • A. Elastic Inference accelerates inference performance by attaching GPU-powered acceleration, not primarily for high availability.
  • C. SageMaker Model Monitor is for detecting data and model quality drift, not for high availability.
  • D. SageMaker Pipelines are for orchestrating ML workflows, not for ensuring real-time endpoint availability.

SageMaker Multi-AZ Endpoint

An Amazon SageMaker endpoint configuration that deploys model instances across multiple AWS Availability Zones to ensure high availability and fault tolerance.

  • Increases resilience against AZ failures.
  • Distributes traffic across zones.
  • Crucial for critical, high-availability ML applications.

Memory trick: Multiple Zones, Multiple Guards.

More Machine Learning Implementation and Operations questions