Amazon Kinesis
Streaming data service family
The umbrella name for AWS's real-time streaming data services: Kinesis Data Streams, Kinesis Video Streams (Firehose and Managed Service for Apache Flink were formerly Kinesis-branded).
- Cost$$ Medium
- Difficulty●●●●●
- Costs money when idle?No
- Serverless
- In examsSAA · DVA
Think of it asLike a factory's conveyor system: different belts for different kinds of goods.
When to use it
✓ Use it when
You need to collect and process continuous data (logs, clickstreams, IoT, video).
✗ Avoid it when
Periodic batch processing → Glue / EMR.
Common uses
- Real-time clickstream collection
- Load logs into a data lake
- Analyze video from cameras
What does it cost?
- Pricing model: Depends on the sub-service: per shard-hour, per GB of data, or by processing time.
When the question says collect and process real-time streaming data (clickstream, IoT, logs), think Kinesis; for loading into S3 or Redshift with no code, Data Firehose; for stream analytics with SQL or Flink, Managed Service for Apache Flink; for video, Kinesis Video Streams; for decoupling components, SQS.
Frequently asked questions
What services are in the Kinesis family and when do I use each?
Kinesis Data Streams ingests and stores streams for many consumers to process in real time. Data Firehose delivers streams to S3, Redshift or OpenSearch with no code. Managed Service for Apache Flink runs stream analytics with SQL or Flink. Kinesis Video Streams ingests video from cameras and devices. The old name Kinesis Data Analytics is now Flink.
How is Kinesis different from SQS?
SQS is a queue that decouples components: each message is typically processed by one consumer and then deleted, with no replay. Kinesis Data Streams keeps records for the retention period, multiple consumers read independently and can replay, and ordering is kept per partition key. For decoupling services and per-message retries, SQS; for analytics streams with multiple consumers, Kinesis.
How is Kinesis different from Amazon MSK?
Both are streaming platforms. Kinesis is an AWS-native service that is simpler, needs less operation and integrates deeply with AWS. MSK is managed Apache Kafka, a fit when you already have Kafka applications or need the standard Kafka ecosystem and APIs, at the cost of understanding and configuring more.
