AWS Certified Data Engineer Associate DEA-C01 Practice Question

A company is building a centralized data lake on Amazon S3 and registers the datasets as governed tables in AWS Lake Formation. Business analysts query the data with Amazon Athena. New business units will frequently introduce additional attributes to the dataset. The solution must automatically expose the new columns while allowing fine-grained, column-level permissions through Lake Formation tag-based access control, and it should require the least ongoing administration. Which table design best meets these requirements?

Store the data in Apache Parquet files and schedule an AWS Glue crawler to update the table's schema. This approach automatically adds new columns to the AWS Glue Data Catalog, enabling them to be secured using Lake Formation tag-based permissions.
Store the data in CSV files partitioned by business unit and manually run ALTER TABLE ADD COLUMN statements whenever new attributes appear.
Create a separate governed table for each business unit, each with its own schema, and combine the tables through Athena views for company-wide reporting.
Store the data as compressed JSON blobs in a single column and let analysts extract fields at query time with Athena JSON functions.

Report Issue

Answer Description

Storing the data in Apache Parquet and scheduling an AWS Glue crawler provides an automated way to handle schema evolution. The crawler detects new columns and updates the table definition in the AWS Glue Data Catalog, making them available for querying via Athena. This columnar design is essential for applying fine-grained, column-level permissions using Lake Formation's tag-based access control. While tagging new columns is a separate administrative step, this architecture is the only option presented that automates schema discovery while supporting the required security model with minimal overhead. Using a columnar format like Parquet also reduces query cost and latency.

Storing JSON inside a single column prevents column-level tagging, because Lake Formation cannot see the nested fields, thus failing to meet the requirement for fine-grained permissions. Manually issuing ALTER TABLE statements for CSV data introduces unnecessary operational overhead and lacks the required automation. Creating separate tables for each business unit multiplies administrative tasks and complicates cross-unit analytics.

Ask Bash

Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.