AWS Certified Data Engineer Associate DEA-C01 Practice Question

An e-commerce company stores clickstream events as Parquet files in Amazon S3 and exposes them to analysts through an external table in Amazon Redshift Spectrum. Product teams often add new, optional attributes to the event JSON, and analysts need to query these columns as soon as they appear without running manual DDL statements. Which approach best supports this type of schema evolution?

Configure Amazon Redshift Spectrum partition projection on the table to discover new attributes without changing the schema.
Load the files into a Redshift table that contains a single SUPER column and query the new attributes with PartiQL expressions.
Trigger an AWS Lambda function that executes ALTER TABLE ADD COLUMN on the external table whenever a new attribute is detected.
Run an AWS Glue crawler after each ingestion and enable schema merging so new columns are automatically added to the Glue Data Catalog table used by Redshift Spectrum.

Report Issue

Answer Description

Redshift Spectrum relies on the AWS Glue Data Catalog for external table metadata. Configuring an AWS Glue crawler to run after each data delivery and to merge new schema changes automatically adds any newly detected columns to the existing table definition in the catalog. Because Redshift Spectrum reads the updated metadata at query time, analysts can immediately select the new fields with no manual ALTER TABLE.

Using a single SUPER column would require loading data into native Redshift tables and abandoning the current Parquet-based Spectrum workflow. A Lambda function that issues ALTER TABLE commands achieves the same result but re-introduces manual maintenance. Partition projection improves partition discovery, not column evolution, so it does not surface new attributes.

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.