MLS-C01 Question 51
Single answerYou are a Machine Learning Specialist working on a recommendation system. You need to schedule a preprocessing job to clean and prepare data daily and ensure that the job automatically retries in case of failure. The preprocessing script is containerized and stored in Amazon ECR. Which service and configuration should you use to schedule and manage this job effectively?
- A
Amazon SageMaker Processing Jobs with an Amazon EventBridge rule
- B
AWS Batch Jobs with an Amazon EventBridge rule
- C
AWS Glue Workflow with a daily trigger
- D
AWS Step Functions with a Task Timer
Show answer and explanation
Correct answer: B
Explanation
AWS Batch is an ideal solution for running containerized workloads, especially when combined with Amazon EventBridge for scheduling and automated retries. This setup ensures that your preprocessing job is executed daily and can handle failure scenarios without requiring complex manual configurations. Other options like SageMaker Processing Jobs or AWS Glue are more tailored for specific use cases, and Step Functions alone cannot natively handle the scheduling requirement.
- A. Incorrect.
Amazon SageMaker Processing Jobs can preprocess data, but they are not inherently designed to support automated retries or direct scheduling. You would need additional configurations to handle retries, making it less suitable in this scenario.
- B. Correct.
AWS Batch Jobs with an Amazon EventBridge rule is the correct choice, as AWS Batch is well-suited for running containerized workloads like your preprocessing script. EventBridge can schedule the job daily, and AWS Batch provides retry mechanisms to handle failures.
- C. Incorrect.
AWS Glue Workflow can trigger scheduled ETL jobs, but it is specifically designed for data transformation and ETL pipelines, not for running arbitrary containerized preprocessing scripts.
- D. Incorrect.
AWS Step Functions with a Task Timer can coordinate workflows, but it is not natively designed for scheduling recurring jobs. You would still need additional services like EventBridge for scheduling.