MLA-C01 · Question #96
An ML engineer needs to merge and transform data from two sources to retrain an existing ML model. One data source consists of .csv files that are stored in an Amazon S3 bucket. Each .csv file consist
The correct answer is B. Create a weekly AWS Glue job that uses the Apache Spark engine. Use DynamicFrame native. AWS Glue with Apache Spark (Option B) is the best fit because AWS Glue is a fully managed ETL service - it handles cluster provisioning, scaling, and teardown automatically, and its native DynamicFrame API simplifies reading from both S3 (CSV) and Aurora via JDBC connectors. Sche
Question
An ML engineer needs to merge and transform data from two sources to retrain an existing ML model. One data source consists of .csv files that are stored in an Amazon S3 bucket. Each .csv file consists of millions of records. The other data source is an Amazon Aurora DB cluster. The result of the merge process must be written to a second S3 bucket. The ML engineer needs to perform this merge-and-transform task every week. Which solution will meet these requirements with the LEAST operational overhead?
Options
- ACreate a transient Amazon EMR cluster every week. Use the cluster to run an Apache Spark job
- BCreate a weekly AWS Glue job that uses the Apache Spark engine. Use DynamicFrame native
- CCreate an AWS Lambda function that runs Apache Spark code every week to merge and
- DCreate an AWS Batch job that runs Apache Spark code on Amazon EC2 instances every week.
How the community answered
(34 responses)- A26% (9)
- B56% (19)
- C12% (4)
- D6% (2)
Explanation
AWS Glue with Apache Spark (Option B) is the best fit because AWS Glue is a fully managed ETL service - it handles cluster provisioning, scaling, and teardown automatically, and its native DynamicFrame API simplifies reading from both S3 (CSV) and Aurora via JDBC connectors. Scheduling it weekly is built-in via Glue triggers or EventBridge, requiring virtually no infrastructure management.
Why the distractors fall short:
- A (Transient EMR): EMR can run Spark jobs, but you must provision, configure, monitor, and terminate the cluster yourself each week - significantly more operational overhead than Glue.
- C (Lambda): Lambda has a 15-minute execution limit and memory constraints that make it unsuitable for processing millions of records from large CSV files.
- D (AWS Batch on EC2): Batch requires managing compute environments and job definitions on EC2 instances, adding infrastructure overhead that Glue abstracts away entirely.
Memory tip: Think of AWS Glue as "serverless Spark for ETL." Whenever an exam question mentions scheduled, recurring data transformation between AWS sources (S3, RDS/Aurora, Redshift) with minimal ops overhead, Glue is almost always the answer over EMR, Lambda, or Batch.
Topics
Community Discussion
No community discussion yet for this question.