Practice exam · AWS
AWS Certified Data Engineer - Associate (DEA-C01) practice exam
Short answer
The AWS Certified Data Engineer - Associate (DEA-C01) practice exam has 50 questions in 130 minutes, with a 72% passing mark, as configured here. The fastest way to know whether you are ready is to sit a full exam on the clock and read your score by domain, not only the total.
Domains covered
- Data Ingestion and Transformation
- Data Operations and Support
- Data Security and Governance
- Data Store Management
Sample questions, with explanations
Three questions from the bank, open to read.
A data team wants to monitor a critical Step Functions data pipeline so they are alerted when executions fail and when the daily run does not start by 02:00 UTC (a missed schedule). They prefer native, low-operational-overhead monitoring. Which combination of services best satisfies both the failure alert and the missed-run alert?
- Enable AWS CloudTrail data events on the state machine and parse them with Athena once per day to find failures.
- Configure the state machine's Catch blocks to write a row to DynamoDB on failure and have someone check the table periodically.
- Use the ExecutionsFailed CloudWatch metric with an alarm to SNS for failures, and an EventBridge Scheduler trigger plus a CloudWatch alarm on ExecutionsStarted (treating missing data as breaching) for missed runs. ✓ (correct)
- Rely solely on reading CloudWatch Logs Insights queries manually each morning to detect both failures and missed runs.
Why this is correct
Using the ExecutionsFailed metric with an SNS alarm plus an ExecutionsStarted alarm that treats missing data as breaching is correct: Step Functions publishes ExecutionsFailed and ExecutionsStarted CloudWatch metrics natively, so a metric alarm to SNS gives immediate failure notification, and an alarm that treats no-data-as-breaching over the expected window detects a missed start without custom code. Manually reading Logs Insights each morning is wrong because it is not automated alerting, adds delay, and depends on a human remembering to run queries. Enabling CloudTrail data events and parsing with Athena daily is wrong because CloudTrail is for API audit, the daily batch parse adds latency, and it does not provide proactive alarms on failures or missed runs. Writing failure rows to DynamoDB and checking periodically is wrong because it requires custom polling, only captures failures (not missed starts), and provides no proactive notification.
A company stores most analytical data in Amazon S3 but keeps current order status in an Amazon Aurora PostgreSQL OLTP database. Analysts need a single Athena query that joins historical S3 data with live order rows in Aurora, without copying the Aurora data into S3 first. Which Athena feature enables this, and what is the key operational consideration?
- Athena partition projection, which can read directly from RDS endpoints when the table is registered in Glue
- Athena CTAS, which transparently federates to relational databases at query time
- Redshift Spectrum, because Athena cannot query external relational sources at all
- Athena Federated Query with the Aurora/JDBC connector (a Lambda function); the Lambda concurrency and source database load must be monitored because each query invokes the connector ✓ (correct)
Why this is correct
Athena Federated Query uses data source connectors implemented as Lambda functions to read from non-S3 sources like Aurora at query time, enabling cross-source joins without copying data; because the connector Lambda issues live queries against Aurora, you must watch Lambda concurrency limits and the load placed on the OLTP database. 'Partition projection reading from RDS' confuses two unrelated features: partition projection is an S3 partition-pruning optimization and cannot read relational endpoints. 'CTAS transparently federates' is wrong because CTAS writes query output to S3 and does not itself provide source connectivity; federation is a separate mechanism. 'Redshift Spectrum because Athena cannot query relational sources' is false on both counts: Athena CAN query relational sources via federation, and Redshift Spectrum specifically queries S3 data from Redshift, not arbitrary live RDS tables.
A startup needs to load a vendor's daily 50 GB JSON export from Amazon S3 directly into Amazon Redshift for analytics. The data needs only light type casting, no complex multi-source joins, and the team wants the fastest, most cost-efficient bulk load using Redshift's native parallelism without provisioning a separate ETL cluster. Which ingestion method should they use?
- Stream the file line-by-line through Amazon Kinesis Data Streams into Redshift.
- Use an AWS Glue Spark job to read the JSON and write each row to Redshift via individual INSERT statements.
- Use the Redshift COPY command to bulk load the JSON files from S3 in parallel across all slices. ✓ (correct)
- Use Amazon Data Firehose to deliver the file into Redshift as a continuous stream.
Why this is correct
The Redshift COPY command is correct because it is the native, highest-throughput bulk-load mechanism, reading files from S3 in parallel across all compute slices with optional simple type handling, and it requires no separate ETL cluster, making it the fastest and most cost-efficient option for this straightforward load. An AWS Glue Spark job issuing per-row INSERT statements is wrong because row-by-row inserts bypass Redshift's parallel load path and are dramatically slower and more expensive, and spinning up Spark is unnecessary for light casting. Streaming the file through Kinesis Data Streams is wrong because the source is a single bounded daily file, and forcing it through a streaming service adds shards, ordering, and consumer complexity with no benefit for a bulk load. Using Amazon Data Firehose is wrong because Firehose is built for continuous streaming delivery and micro-batches into Redshift via COPY under the hood, which is slower and less direct than simply issuing one COPY for a known daily file.
Take the full practice exam
- 50 questions drawn from the bank, with a 130-minute timer
- Score by domain and a review of every question
- An AI tutor to talk through each question you miss
- What you miss feeds your study plan
Independent, unofficial practice exam. AWS and the exam name are trademarks of their owners; UpSkillScore is not affiliated with them. Format and passing mark are the exam's configuration in UpSkillScore, not an official statement.
Frequently asked questions
How many questions does the AWS Certified Data Engineer - Associate (DEA-C01) practice exam have?
It has 50 questions and a 130-minute limit. The passing mark of the practice exam is 72%.
Is this the official AWS Certified Data Engineer - Associate (DEA-C01) exam?
No. It is an independent practice exam built to match the structure of the real one. It does not replace the official exam.
Do I need an account to take it?
You can read the structure and sample questions without an account. Taking the full timed exam, scoring by domain and the AI tutor need one.