The AWS Data Service Comparisons That Decide DEA-C01 Questions
← Back to all guides

DEA-C01

The AWS Data Service Comparisons That Decide DEA-C01 Questions

Glue or EMR? Kinesis or MSK? Redshift or Athena? The data-service trade-offs that decide DEA-C01 questions, explained by use case.

A

Aziz FaridAWS certification prep · DEA-C01

DEA-C01 questions often describe a data requirement and ask you to choose the AWS service or architecture that fits. The useful skill is comparing what each service is designed to do, what it does not do, and which trade-off matters in the scenario.

1. Kinesis Data Streams vs. Data Firehose vs. Amazon MSK

Kinesis Data Streams is for custom real-time streaming applications. You control consumers and processing, and can replay records within the stream’s retention period. Choose it when you need application-level processing, multiple consumers, or control over how records are handled.

Data Firehose is a managed delivery service for streaming data to supported destinations, with buffering and optional transformations. Choose it when the requirement is to deliver streaming data with less infrastructure to manage—not when you need a custom consumer to process every record.

Amazon MSK is a managed Apache Kafka service. Consider it when the scenario specifically needs Kafka compatibility, Kafka APIs, or an existing Kafka ecosystem. Do not assume MSK, Kinesis Data Streams, and Firehose are interchangeable just because all can appear in streaming architectures.

2. AWS Glue vs. Amazon EMR

AWS Glue is a serverless data integration service with managed ETL capabilities and catalog integration. It is a strong fit for managed, scheduled or event-driven ETL when you want AWS to handle much of the infrastructure.

Amazon EMR runs big-data frameworks such as Apache Spark with more choice and control over the runtime and cluster configuration. Consider it when the workload needs framework flexibility, specific configuration, or control that a managed Glue job does not provide. Compare operational effort as well as runtime cost.

3. Amazon Athena vs. Amazon Redshift

Amazon Athena runs SQL queries directly against data in sources such as Amazon S3, without first loading that data into a warehouse. It is useful for ad hoc analysis and data-lake queries; pay attention to scanned data, partitioning, and file formats.

Amazon Redshift is a data warehouse for analytical workloads that benefit from modeled data, repeated queries, and warehouse-oriented performance features. Choose based on query patterns, concurrency, data layout, performance needs, and total cost—not simply because both services use SQL.

4. Amazon S3 data lake vs. a data warehouse

Amazon S3 provides object storage for raw, staged, and curated data. It is a common foundation for a data lake, but storing files in S3 does not automatically provide a catalog, data quality, transactional table behavior, or governed access.

Amazon Redshift provides warehouse capabilities for analytical querying. A scenario may use both: S3 for durable, flexible data storage and Redshift for curated analytical workloads. Look for the required access pattern and governance model before choosing.

5. AWS Glue Data Catalog vs. AWS Lake Formation

AWS Glue Data Catalog stores technical metadata such as table definitions, schemas, and locations so data can be discovered and used by analytics services.

AWS Lake Formation helps centrally define and enforce fine-grained permissions for data-lake resources. A catalog makes data discoverable; governance controls who can access it. Many designs use both rather than treating them as substitutes.

6. AWS Lambda vs. AWS Step Functions

AWS Lambda runs code in response to events and is suitable for an individual processing step. AWS Step Functions coordinates multi-step workflows, including sequencing, branching, retries, and error handling. Use Lambda for the work a function performs and Step Functions when the scenario calls for visible workflow orchestration.

7. AWS DMS vs. AWS Glue

AWS Database Migration Service (AWS DMS) is designed to migrate databases and replicate ongoing changes, including change data capture for supported sources and targets. AWS Glue is designed for data integration and transformation. If the requirement emphasizes database migration or ongoing replication, evaluate DMS; if it emphasizes transforming and preparing data, evaluate Glue.

8. Apache Iceberg and open table formats

Do not treat an open table format as just another file extension. Apache Iceberg adds table-level metadata and capabilities for managing large analytic tables on data lakes. For DEA-C01, understand why table formats, partition evolution, schema evolution, and reliable table operations matter. Check the current AWS exam guide for the exact skills and service scope.

9. Batch vs. streaming—and ETL vs. ELT

Batch processes bounded data on a schedule or in discrete runs. Streaming processes events as they arrive and is justified when the required latency or continuous processing demands it. Streaming introduces additional design considerations such as ordering, replay, late-arriving events, and consumer behavior.

ETL transforms data before loading it into the target. ELT loads data first and transforms it in the target or query environment. The correct choice depends on data quality, governance, target capabilities, performance, and cost—not on a universal rule that one pattern is better.

DEA-C01 decision signals

  • Near real-time: Compare streaming services and confirm latency, ordering, replay, and consumer requirements.
  • Low operational overhead: Consider managed/serverless services, but check whether their limits fit the workload.
  • Custom processing or Kafka compatibility: Distinguish Kinesis Data Streams from MSK and managed delivery services.
  • Ad hoc SQL on S3: Consider Athena; repeated warehouse analytics may point to Redshift.
  • Migration and change capture: Evaluate DMS; transformation pipelines may point to Glue or EMR.
  • Governed data sharing: Think about catalog, permissions, encryption, lineage, privacy, and audit—not just storage.
  • Cost and performance: Inspect partitioning, file format, compression, scanned data, compute time, storage lifecycle, and data movement together.

These comparisons support the four official DEA-C01 domains; they are not a prediction that a specific service or question will appear. Start with the DEA-C01 exam domains and topics guide, then use the DEA-C01 study plan to organize your preparation. For a baseline, try the free 25-question diagnostic.

Ready to test what you know?

Try a free 25-question AWS diagnostic — no sign-up, instant results and domain-level feedback.

Start a free diagnostic →

Limited-time offer · ends in -- Days : -- Hrs : -- Min : -- Sec
Get Instant Access