A company runs a batch inference job every night on a large dataset. The job takes 2 hours to complete on a single ml.c5.4xlarge instance. The company wants to reduce the cost of this job. The job is fault-tolerant and can be interrupted. What should the company do?

Move the job to a SageMaker real-time endpoint and invoke it with a script.

Use a smaller instance type, such as ml.c5.xlarge, and increase the job timeout.

Use Spot Instances for the training job.

Use a larger instance type, such as ml.c5.9xlarge, to complete the job faster.

A startup runs a small ML service that performs real-time inference for a mobile app. The traffic is highly variable, with occasional spikes. The team wants to minimize operational overhead and cost, and they do not want to manage servers. What should they use?

Amazon ECS on EC2 with a load balancer.

A SageMaker real-time endpoint with automatic scaling.

SageMaker Serverless Inference.

AWS Lambda with a container image for the model.

AWSCertified Machine Learning Engineer - Associate

Domain 4Objective 2

Task 4.2: Monitor and Optimize Infrastructure and Costs MLA-C01 Practice Questions (Page 5)

Part of the Content Domain 4: ML Solution Monitoring, Maintenance, and Security domain, which accounts for 24% of the MLA-C01 exam. AWS does not publish an official question count, but from its 130-minute exam (~50–85 total, ~12–20 in this domain), expect 4–7 from this objective — we provide 27 practice questions to prepare you well beyond it. (estimate)

27questions here

6free pages

8concepts

24%of the exam

Questions 21–25

21foundation · easy
Which EC2 purchasing option is most cost-effective for a fault-tolerant ML training job that can be interrupted and resumed?
Select an answer first
22foundation · easy
Which AWS service is the primary source for collecting and storing infrastructure metrics such as CPU utilization, network throughput, and disk I/O for Amazon EC2 instances running ML workloads?
Select an answer first
23application · medium
A company runs a batch inference job every night on a large dataset. The job takes 2 hours to complete on a single ml.c5.4xlarge instance. The company wants to reduce the cost of this job. The job is fault-tolerant and can be interrupted. What should the company do?
Select an answer first
24application · medium
A team is troubleshooting a SageMaker real-time endpoint that is experiencing high latency. The team suspects the instance is running out of memory, causing swapping. Which CloudWatch metric should the team monitor to confirm this?
Select an answer first
25application · medium
A startup runs a small ML service that performs real-time inference for a mobile app. The traffic is highly variable, with occasional spikes. The team wants to minimize operational overhead and cost, and they do not want to manage servers. What should they use?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to MLA-C01

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by AWS. “MLA-C01” is a trademark of its owner, used for identification only.

Task 4.2: Monitor and Optimize Infrastructure and Costs MLA-C01 Practice Questions (Page 5)

Questions 21–25

Which EC2 purchasing option is most cost-effective for a fault-tolerant ML training job that can be interrupted and resumed?

Which AWS service is the primary source for collecting and storing infrastructure metrics such as CPU utilization, network throughput, and disk I/O for Amazon EC2 instances running ML workloads?

A company runs a batch inference job every night on a large dataset. The job takes 2 hours to complete on a single ml.c5.4xlarge instance. The company wants to reduce the cost of this job. The job is fault-tolerant and can be interrupted. What should the company do?

A team is troubleshooting a SageMaker real-time endpoint that is experiencing high latency. The team suspects the instance is running out of memory, causing swapping. Which CloudWatch metric should the team monitor to confirm this?

A startup runs a small ML service that performs real-time inference for a mobile app. The traffic is highly variable, with occasional spikes. The team wants to minimize operational overhead and cost, and they do not want to manage servers. What should they use?

Keep studying — Content Domain 4: ML Solution Monitoring, Maintenance, and Security