DEA-C01 sample questions with answers
The other 221 questions are in the practice tool, with answers revealed as you go.
Start 20 questionsData ingestion and transformation
- An e-commerce platform writes daily order events as JSON files to S3. Analysts want the data converted to Parquet and organized into folders by date to reduce scan costs in Athena. Which approach is most suitable?
- A data pipeline must wait for a new file to land in S3, start a Glue job, retry that job on failure, and send a notification when it completes. Which service coordinates these steps most suitably?
- A team runs Apache Airflow on Amazon MWAA and wants DAG tasks to create and monitor S3 objects and start Glue jobs using AWS-aware operators. What should the team do?
- Before loading incoming records into the warehouse, a team needs managed rule checks for completeness, uniqueness and value ranges on tables registered in the Data Catalog. Which AWS service provides these checks?
- A platform team runs one Kinesis data stream with 12 shards that feeds a fraud model, a live dashboard, and an archive job. Each reader must get dedicated read bandwidth so that none of them slows the others. Under default quotas, how many registered consumers can use enhanced fan-out on this single stream?
- A data engineer's AWS Glue job runs in us-east-1 and must transform Parquet files stored in an S3 bucket located in eu-west-1. Which statement describes the correct approach?
- A nightly AWS Glue Spark job fails with ExecutorLostFailure after running for about 40 minutes. The driver logs show one executor's container being killed, and the failing stage processes a single partition that holds most of the data. Which cause is MOST likely?
- A team creates a new AWS Glue batch ETL job from the console and leaves the timeout field at its default. After how long will Glue stop a run that is still executing?
- An e-commerce company needs the orders table in Amazon RDS for PostgreSQL reflected in an Amazon Redshift warehouse within minutes, including updates and deletes made at the source after the initial load. Which approach fits best?
- A Glue job reads JSON event files in which one field sometimes holds a string and sometimes a number, which breaks a fixed schema. Which Glue abstraction handles this per-record variation without failing the job?
- A nightly pipeline writes curated sales tables. The team wants the Spark job itself to stop and prevent publication when a null-rate or uniqueness constraint fails. Which approach meets that need?
Data store management
- A social platform needs to store and traverse a network of users, follows, and shared interests, with hundreds of billions of edges that must be queried by multi-hop relationship patterns. Which managed service fits best?
- A university portal stores students, courses and enrollments in one DynamoDB table. The app must list all courses for a student and all students in a course, with each query served by key lookups rather than scans. Which data model should the team use?
- A data team writes clickstream events as Parquet files in S3 and queries them interactively with Athena. They want a compression codec that keeps decompression fast so dashboards stay responsive. Which codec should they choose?
- A team wants to run standard SQL against a folder of CSV files stored in S3 for ad hoc questions, without loading the files into a database or provisioning a cluster. Which service is the best fit?
- A team's raw clickstream files in S3 gain new columns every few weeks, and Athena queries break because the catalog table still shows the old layout. Which setup keeps the catalog aligned without manual table edits?
- A multiplayer game keeps short-lived player session records keyed by a session identifier. The records are small, traffic varies by time of day, and the service must answer reads and writes in single-digit milliseconds as the player base grows. Which data store fits best?
- A team must lift an existing Apache Cassandra workload to AWS. The application uses CQL, expects tunable consistency, must keep single-digit millisecond reads, and needs the data replicated to a second Region for local access. Which service is the closest fit?
- A platform team must change the storage engine and several parameters on a production RDS for MySQL instance. They want a managed staging copy that receives the changes, can be validated with production-like traffic, and supports a fast switchover with rollback. Which capability should they use?
- A high-transaction OLTP database on RDS for MySQL needs a sustained 40,000 IOPS with consistent single-digit millisecond latency, independent of how much data is stored. Which volume type should be chosen?
Data operations and support
- A product team runs MySQL on Amazon RDS and wants to restore the database to how it looked 3 hours ago after a faulty migration. The backup feature must be native to RDS and need no extra setup. Which option meets this?
- An engineering manager wants automated checks that flag idle or underused resources across the account, including warehouse and storage services, plus suggestions for lowering spend. Which service should the manager review?
- A streaming ETL pipeline feeds a customer-facing dashboard. If an entire AWS Region becomes unreachable, the pipeline must keep ingesting and serving within minutes rather than waiting for that Region to recover. Which design meets the requirement?
- A finance team's critical Redshift reporting cluster must return to service with the smallest possible recovery time after a hardware or Availability Zone failure. Which option gives the lowest RTO?
- A Lambda function writes plain-text log lines such as 'order=1182 status=FAILED region=eu-west-1' to a CloudWatch log group. An engineer wants the order ID and status to appear as queryable fields in Logs Insights. Which command pulls those values out of the message text?
- An Athena query fails after running for several minutes, and the team needs stage-level timing and data volumes to find the bottleneck. What should they do first?
- An Aurora PostgreSQL cluster serves 500 application servers. The team needs connection pooling, IAM-based database credentials instead of embedded passwords, faster recovery when the writer fails, and a cache for repeated read queries. Which combination covers all four?
- CloudWatch shows ReadLatency on an RDS for MySQL instance climbing from about 2 ms to 40 ms while CPU stays moderate. Queries that use indexes are also getting slower. Which cause should be investigated first?
Data security and governance
- A bank's QA team needs realistic customer tables for load testing, but production names, account numbers and emails cannot leave the production boundary. Which technique should the data team apply before copying the data?
- A partner firm's consultant needs temporary read access to one S3 prefix in your account. Your policy requires no long-lived credentials and must limit how the grant can be misused by others. Which design meets these needs?
- A security team wants a managed analyzer that flags S3 buckets or IAM roles that grant access to principals outside the organization's account or organization boundary. Which service provides this?
- A compliance analyst is triaging Amazon Macie sensitive data findings and wants to know the severity scale used to rank them. Which set is correct?