AWS Certified Data Engineer Associate DEA-C01 Practice Question

Staging the exported tables in Amazon S3 keeps storage costs low and accommodates incremental uploads over a constrained network. Writing the data as compressed Parquet further reduces transfer size and improves scan efficiency. By defining external tables, Amazon Redshift Spectrum lets analysts run familiar SQL while a minimally sized Redshift cluster stores only metadata, so compute cost is limited until the final cutover. Moving the data into DynamoDB would force a non-analytic store that lacks full SQL support, Aurora would incur higher storage and instance costs for 20 TB, and maintaining an EMR cluster for several months adds unnecessary operational and infrastructure expense.