Move and transform data
Compare ingestion and orchestration approaches for batch, streaming and changing source data.
DEA-C01 · Associate
AWS Certified Data Engineer – Associate
Build better judgment across the AWS data pipeline—from ingestion to dependable delivery.
Your free account includes a full DEA-C01 practice exam.
ExamNova practice
Your first session
- Questions
- 65
- Time limit
- 130 min
- Coverage
- 4 domains
Practise, review your answers and see where to focus next.
Preview a questionIs DEA-C01 your next step?
For data engineers implementing and operating AWS pipelines and data stores. The exam brings together data movement, transformation, reliability, access control and cost/performance decisions.
Choose the data store
Match storage, cataloguing and processing choices to the access pattern and performance requirements.
Operate a trustworthy pipeline
Work through data quality, monitoring, troubleshooting, permissions and governance.
What you’ll study
Explore the domains and topics in your ExamNova study path.
Data Ingestion and Transformation34% practice balance
Explore Data Ingestion and Transformation
Perform Data Ingestion
- Batch Performance and Connector Tuning
- Batch Triggers and Scheduling
- Hybrid and Partner Data Ingestion
- Source Connectivity and Transfer
- Idempotent Stream Processing
- Producer and Consumer Patterns
- Stateful Stream Processing
- Stream Delivery and Routing
- Streaming Reliability and Replay
- Streaming Source Services
Transform and Process Data
- Glue and EMR Batch Processing
- Data Profiling and Preparation
- Deduplication and Enrichment
- Schema Discovery and Mapping
- Transformation Troubleshooting
- Columnar File Formats
- Compression and Partitioning
- Format Selection Tradeoffs
- Table Formats and Catalog Integration
- Query and Partition Optimization
- Storage Format Optimization
Orchestrate Data Pipelines
- Event-Driven Pipeline Triggers
- Job Dependencies and Retries
- Operational Notifications and Alerts
- Serverless Pipeline Patterns
- Pipeline Resilience Planning
- High Availability Patterns
Apply Programming Concepts
- Serverless Performance Tuning
Data Store Management26% practice balance
Choose a Data Store
- Open Table Formats
- S3 Data Lake Layout
- Redshift Architecture
- Spectrum and Federated Queries
- Warehouse Monitoring and Scaling
- Key-Value and Document Access Patterns
- Scaling and Throughput Controls
- Specialized NoSQL Services
- Availability and Maintenance
- Connectivity and Migration
- Query Access and Federation
- RDS and Aurora Selection
- Relational Analytics Read Offload
Understand Data Cataloging Systems
- Catalog and Crawler Integration
- Business Metadata and Ownership
- Schema Discovery and Partition Sync
- Technical Data Cataloging
Manage the Lifecycle of Data
- Lifecycle and Storage Optimization
- Data Loading and Unloading
- Backup, TTL, and Global Tables
- Storage Lifecycle Savings
- Backup and Restore Strategy
- Cross-Region Replication
- Recovery Objectives and Testing
Design Data Models and Schema Evolution
- Schema Conversion and Evolution
- Schema Compatibility and Semantic Governance
- Distribution, Sort, and Compression
- DynamoDB Data Modeling
- Data Lineage Tracking
Data Operations and Support22% practice balance
Explore Data Operations and Support
Automate Data Processing by Using AWS Services
- Workflow Orchestration Services
Analyze Data by Using AWS Services
- Cost-Performance Tradeoffs
- Compute Right-Sizing
- Service-Specific Pricing Tradeoffs
- Redshift Query Performance Diagnosis
- Data Visualization and Business Intelligence
Maintain and Monitor Data Pipelines
- Budget Controls and Anomaly Detection
- Cost Visibility and Allocation
- CloudTrail and Audit Visibility
- CloudWatch Logs and Alarms
- Notification and Escalation Patterns
- Pipeline Health Metrics
- Glue and EMR Job Tuning
- Job Failure Diagnosis
- Log Analysis Workflows
- Performance Troubleshooting
- Retry and Error Handling
Ensure Data Quality
- Quality Rules and Validation
- Data Quality Monitoring
- Data Consistency Investigation
Data Security and Governance18% practice balance
Explore Data Security and Governance
Apply Authentication Mechanisms
- IAM Roles and Policies for Data Services
- Private Endpoint Access
- Secrets and Credential Protection
Apply Authorization Mechanisms
- Access Review and Governance
- Database User and Group Access
- Lake Formation Permissions
- Least Privilege Design
- Lake Formation Governance
Ensure Data Encryption and Masking
- Cross-Account Encryption
- Encryption at REST
- Encryption in Transit
- KMS Key Management
- Anonymization and Tokenization
- Compliance-Driven Protection
- Data Masking Patterns
Prepare Logs for Audit
- CloudTrail Audit Logging
- Compliance Evidence Collection
- Log Retention and Query
Understand Data Privacy and Governance
- Config and Resource Change Tracking
- Data Sharing and Governance
- PII Discovery
- Privacy Controls in Pipelines
Percentages show ExamNova’s practice balance. Consult the official guide for the vendor’s current exam outline.
Try focused practice
Explore a topic with sample questions and explanations.
Try a question before you start.
Read the scenario. Consider your answer, then reveal the reasoning.
Trace where a record enters, changes and is consumed. Pipeline questions often turn on freshness, failure recovery or access needs.
A Lambda consumer processes Kinesis records but falls behind during traffic spikes. The function duration is high, memory usage is near the configured limit, and throttles appear during the spike. Which TWO changes should the engineer evaluate first?
- Move the stream to S3 Glacier Deep Archive.
- Review reserved concurrency and event source parallelization so the consumer can scale appropriately.
- Increase Lambda memory to improve available CPU and test the effect on duration.
- Disable CloudWatch metrics to reduce Lambda overhead.
- Reduce Kinesis retention to one hour so the backlog disappears faster.
Show answer and explanations
Move the stream to S3 Glacier Deep Archive. — Kinesis records are not processed by moving the stream to S3 archival storage.
Review reserved concurrency and event source parallelization so the consumer can scale appropriately. — Reserved concurrency, event source settings, and stream scaling determine how much parallel processing the consumer can use.
Increase Lambda memory to improve available CPU and test the effect on duration. — For Lambda, increasing memory also increases CPU allocation and can reduce processing duration when the function is resource constrained.
Disable CloudWatch metrics to reduce Lambda overhead. — Disabling observability makes tuning harder and does not solve throttling.
Reduce Kinesis retention to one hour so the backlog disappears faster. — Shorter retention does not improve processing and may cause unrecovered records to expire.
A clearer way to prepare.
One place to practise, understand your results and plan the next session.
Practise with purpose
Start with your free exam. Explore Exam, Endless and Custom practice modes as you build your study routine.
Understand the answer
Review the reasoning behind your answers and return to the decisions that need another look.
Find your next focus
Use domain-level performance and readiness to see strengths, gaps and areas you have yet to assess.
Put it into practice.
Explore hands-on tasks connected to DEA-C01. These are real Labs from the catalogue.
Lab
Turn clickstream events into user sessions
Write one read-only query in session-audit.sql against web_events. Return exactly user_id, session_number, session_start, session_end, and event_count, with one row per session. event_time contains UTC epoch seconds. Order each user's events by event_time and use event_id to break ties. Number sessions from 1 separately for each user. Start a new session only when the gap from that user's previous event exceeds 30 minutes; exactly 30 minutes remains in the same session. Retain every event and order the result by user_id, then session_number.
hard · About 20 min
LabKeep the nightly transform on local time
Repair pipeline-schedule.yaml so a single enabled schedule named orders-nightly-local starts the Glue job at 02:15 Australia/Sydney every day, with no delivery window. Start orders-nightly in incremental run mode, retry twice within an hour before falling back to ScheduleDLQ, and leave ScheduleDLQ itself unchanged. The execution role should be assumable only by EventBridge Scheduler and hold no permission beyond starting that job and writing to that queue.
medium · About 18 min
Make DEA-C01 your next step.
Create your free account. Start practising.
Start Free