AWS Certified Data Engineer - Associate Practice Test
The AWS Certified Data Engineer - Associate (DEA-C01) concerns building and operating data pipelines and data stores on AWS. The work includes moving data between sources and destinations, preparing datasets for analysis, and choosing storage and processing approaches that balance query performance with cost.
Exam overview
About the AWS Certified Data Engineer - Associate Exam
Data engineering decisions often begin with the shape and arrival pattern of data: batch files, database change events, or continuous streams. Practice topics include selecting ingestion and replication services, processing S3 data with managed Spark, and organizing objects so analytics engines scan fewer bytes. Other scenarios focus on cataloging schemas for discovery across services and choosing a low-latency key-value store for application access. The professional task is to match a workload’s latency, scale, access pattern, and operational constraints to suitable AWS components, then consider how file layout and processing choices affect reliability, performance, and spend. Questions can help exercise service selection and trade-off reasoning across these pipeline stages.
Sample Practice Questions
Review a selection from this exam's published US Practice Tests practice bank.
A company has an Amazon S3 data lake with many tiny files that are slowing down analytics queries. What is the best general practice to improve scan efficiency?
Show answer
Correct answer: AConsolidating many small files into fewer larger files usually reduces object overhead and improves query performance for analytical scans.
A company needs to ingest change events from a MySQL database into a data lake on Amazon S3 with low latency. Which approach is most appropriate?
Show answer
Correct answer: BChange data capture, or CDC, captures inserts, updates, and deletes as they happen or near real time, which is appropriate for keeping a data lake current. Nightly full exports are slower and more costly, SQS is a message queue rather than a data replication mechanism, and CloudFront invalidations are unrelated.
Which AWS service is commonly used to ingest streaming data and allow multiple consumers to read from the stream independently?
Show answer
Correct answer: AAmazon Kinesis Data Streams is designed for near-real-time streaming ingestion with multiple consumers able to process the data.
A table in Amazon S3 is queried repeatedly by analysts. To reduce query cost and improve performance, which data layout change is generally the best practice?
Show answer
Correct answer: CPartitioning by commonly filtered columns can reduce the amount of data scanned and improve performance because queries can skip irrelevant partitions.
A company needs to transform raw data in Amazon S3 using Spark, but it wants the environment to scale without managing servers and to pay only while the job runs. Which AWS service best matches this need?
Show answer
Correct answer: AAWS Glue provides serverless ETL with Spark-based jobs, so it is a common choice for transforming data in S3 without managing infrastructure continuously.
A data engineer wants a central metadata store so that multiple AWS analytics services can discover table definitions and schemas. Which component is designed for this purpose?
Show answer
Correct answer: BThe AWS Glue Data Catalog is a central metadata repository for datasets, schemas, and table definitions used by services such as Glue, Athena, and EMR. The other services serve different purposes and do not act as a metadata catalog for analytics tables.
A team needs to capture database changes from an on-premises PostgreSQL database and move them continuously into AWS with minimal downtime. Which AWS service is commonly used for this type of migration and replication?
Show answer
Correct answer: AAWS DMS is commonly used for database migration and ongoing change replication, including continuous capture of changes from supported source databases.
How to use this practice test
The free practice bank contains 50 questions, and each session presents up to 20 random questions. Complete a session to see your results, then use missed or uncertain items to identify whether you need to revisit ingestion, S3 analytics layout, Spark processing, metadata catalogs, or data-store selection.
Preparation tips
- For S3 analytics scenarios, compare small-file compaction and columnar, partitioned layouts in terms of scan overhead and bytes queried.
- Distinguish streaming ingestion from database change-data capture, and identify when a replication service is needed to move ongoing database changes.
- Practice matching serverless Spark processing needs to managed job services, including the requirement to avoid persistent cluster administration.
- Trace how a central catalog supports table and schema discovery across analytics services, rather than confusing metadata management with data storage.
- For data-store questions, distinguish key-value access patterns and low-latency application reads from analytical queries over lake data.