Analytics
Amazon Managed Streaming for Apache Kafka (Amazon MSK)
Amazon MSK operates Apache Kafka clusters or serverless Kafka while preserving familiar Kafka APIs. It fits event-heavy GenAI platforms that already depend on Kafka partitions, consumer groups, and ecosystem tooling.
Key points
- MSK supports provisioned and serverless deployment models
- Kafka partitions provide ordering only within a partition
- IAM authentication is available alongside supported Kafka security mechanisms
When to use it
- Distribute document-change events to independent indexing consumers
- Feed durable conversation events into analytics and compliance pipelines
Exam tips
- Prefer MSK for Kafka compatibility and long-lived event streams; use Kinesis for AWS-native stream operations
- Monitor consumer lag and choose keys carefully to avoid hot partitions