Basic Information
| Field | Details |
|---|---|
| English name | Amazon Kinesis Data Streams & Amazon Data Firehose |
| Full name | Amazon Kinesis Data Streams / Amazon Data Firehose |
| Chinese description | 实时数据流与托管式数据交付 |
| Japanese description | リアルタイムデータストリームとマネージドデータ配信 |
| Exam frequency | ⭐⭐⭐⭐ |
| Often confused with | Kinesis Data Streams / Amazon Data Firehose / Amazon MSK |
In one sentence
Data Streams ingests, retains, and replays real-time streams; Data Firehose buffers and delivers them to analytics destinations with less operational work.
Key points
- Data Streams supports multiple consumers, custom processing, and replay; capacity and shard design still matter.
- Data Firehose scales automatically and delivers to S3, Redshift, OpenSearch, and other targets, with optional Lambda transformation.
- Firehose buffers records, so delivery is near real time rather than immediate per record.
Exam focus
- Choose Data Streams for multiple consumers, custom reads, or replay.
- Choose Data Firehose for low-operations delivery to a fixed destination.
Common pitfalls
- Do not choose by product name alone; confirm data type, latency, control, operations, and cost constraints.
Remember
Streams is for streaming and replay; Firehose is for buffering and delivery.
Update: Streams and Firehose
- Kinesis Data Streams is a replayable real-time stream. The partition key selects a shard and records are ordered within that shard; retention can be extended.
- Shared throughput and enhanced fan-out provide different consumer throughput models.
- Amazon Data Firehose is destination-oriented, near-real-time managed delivery with size/time buffering, transformation, compression, and format conversion.
- Redshift delivery stages data in S3 before COPY. Firehose is not a general replay stream for independent consumers.