MLA-C01 Question 426
Select 3You are managing a machine learning workload on AWS that involves multiple EC2 instances, S3 storage buckets, and SageMaker resources. Your team wants to track costs for different components of the project, such as data preprocessing, model training, and inference, to optimize spending. Which of the following steps should you take to enable cost tracking and allocation for this workload?
- A
Use resource tags to categorize EC2 instances, S3 buckets, and SageMaker resources based on the project components they belong to.
- B
Enable AWS Cost Explorer and activate cost allocation tags to track resource costs based on tags.
- C
Manually record the cost of each resource daily in a spreadsheet for tracking.
- D
Set up AWS Budgets to monitor costs for specific tags and notify the team when spending exceeds thresholds.
- E
Configure Amazon CloudWatch to monitor the cost of tagged resources in real time.
Show answer and explanation
Correct answers: A, B, D
Explanation
To track and allocate costs effectively for a machine learning workload, you should use resource tagging to categorize resources, activate cost allocation tags in AWS Cost Explorer, and set up AWS Budgets for monitoring and alerts. These AWS tools are specifically designed for cost tracking and management, eliminating the need for manual processes and providing detailed insights into spending.
- A. Correct.
Correct: Tagging resources with relevant categories (e.g., preprocessing, training) allows you to group and track costs effectively in AWS Cost Management tools.
- B. Correct.
Correct: Activating cost allocation tags and using AWS Cost Explorer provides a detailed breakdown of costs based on tags, enabling better cost tracking.
- C. Incorrect.
Incorrect: Manually recording costs is error-prone and inefficient when AWS provides automated tools for tracking and allocation.
- D. Correct.
Correct: AWS Budgets can be configured to monitor costs for specific tags and send alerts, helping to control spending effectively.
- E. Incorrect.
Incorrect: While CloudWatch is great for monitoring performance and operational metrics, it does not provide detailed cost tracking based on tags.