Get Your CertMLA-C01PlayCompete
‹ All MLA-C01 questions

MLA-C01 Practice Question 88: Scheduled Pipeline Retraining

AWS Certified Machine Learning Engineer – Associate (MLA-C01) · Test 9 · question 8 of 10

Deployment and Orchestrationeasy

An ML team has a SageMaker Pipelines pipeline that retrains a demand model on the latest week of data, evaluates it, and registers it. The pipeline must run automatically every Sunday at 02:00 UTC. The team does not want to manage any servers, write polling code, or keep a notebook running for this.

How should the team schedule the pipeline?

Select one answer
Show answer and explanation

Answer:

  • B. Create an Amazon EventBridge schedule with a weekly cron expression that targets the pipeline and starts a new execution

SageMaker Pipelines is supported as an Amazon EventBridge target, and EventBridge can start a pipeline execution on a schedule. A weekly cron expression for Sunday at 02:00 UTC starts the retraining pipeline with no servers, no polling code, and nothing left running between runs (option B).

- A is wrong because the EC2 instance is a server the team would have to run, patch, and pay for around the clock, just to fire one API call a week. - C is wrong because a notebook instance running a sleep loop is a server the team must keep up, and the schedule dies whenever the notebook restarts. - D is wrong because a looping state machine with a seven-day wait is a fragile stand-in for a scheduler. It drifts from the fixed Sunday time as run times vary, and it stops if one execution fails.

Reference: Schedule Pipeline Runs

Practice Test 9 ›

Study guides for MLA-C01

AWS ML Engineer Associate, MLA-C01 to MLA-C02What changes from MLA-C01 to the MLA-C02 beta of the AWS ML Engineer Associate exam: format, domain weights, new GenAI skills and how to prepare.