MLS-C01 Question 319
Single answerYou are building a machine learning pipeline on AWS that processes data using Amazon SageMaker training jobs. During testing, you notice that your training jobs are being throttled due to service quota limits. What should you do to resolve this issue?
- A
Request a service quota increase for Amazon SageMaker training jobs through the Service Quotas console.
- B
Enable auto-scaling for your SageMaker training jobs.
- C
Switch to a larger SageMaker instance type to avoid hitting service quotas.
- D
Use AWS Trusted Advisor to automatically increase the quota for SageMaker training jobs.
Show answer and explanation
Correct answer: A
Explanation
AWS service quotas are predefined limits for resources and services, such as the number of SageMaker training jobs that can run concurrently. If you encounter throttling due to these limits, the correct approach is to request a quota increase through the Service Quotas console or AWS Support. Other options, such as changing instance types or enabling auto-scaling, do not address service quota limits.
- A. Correct.
This is the correct approach. Service quotas are limits set by AWS and must be increased by submitting a request through the Service Quotas console or AWS Support. This is the proper way to handle throttling related to service quotas.
- B. Incorrect.
Auto-scaling is not relevant in this case. Auto-scaling adjusts the number of resources but does not address service quota limits.
- C. Incorrect.
Changing the instance type does not affect service quotas. Service quotas apply to resources like the number of training jobs, not the instance size.
- D. Incorrect.
AWS Trusted Advisor provides recommendations on cost optimization, security, and performance but does not automatically increase service quotas.