In one sentence
Both handle streaming data, but Data Streams provides a readable, replayable stream while Data Firehose provides managed delivery.
Key points
- Multiple consumers, custom processing, read control, and replay → Data Streams.
- Automatic scaling, buffering, and delivery to S3, Redshift, or OpenSearch → Data Firehose.
- They can appear together: Streams ingests and Firehose delivers.
Exam focus
- Do not interpret Firehose buffering as immediate delivery of every record.
Common pitfalls
- Do not choose by product name alone; confirm data type, latency, control, operations, and cost constraints.
Remember
Need to read and replay the stream? Streams. Just deliver it? Firehose.