Near-Real-Time Analytics with Redshift Streaming Ingestion & Kinesis

Part 8: Hands-on Tutorials · Hands-on Tutorials

Read the full tutorial: Near-Real-Time Analytics with Redshift Streaming Ingestion & Kinesis ↗

Link verified 2026-08-23.

This tutorial streams events through Kinesis Data Streams directly into Redshift using native streaming ingestion — no intermediate S3 landing zone or batch load step — and queries them within seconds of arrival.

It’s the concrete version of the latency call in Should This Be Streaming At All? RT vs NRT Trade-offs & Exactly-Once Semantics, and the Kinesis-to-warehouse path it uses is the same shape as the speed layer in Kappa Architecture: Stream-Only Processing.

flowchart LR
    A[Event producers] --> B[Kinesis Data Streams]
    B --> C[Redshift streaming ingestion - materialized view]
    D[(DynamoDB - reference data)] --> C
    C --> E[NRT SQL queries / dashboards]

---

| ← Previous: Manage Data Transformations with dbt in Amazon Redshift | Next: Data Mesh at Scale with AWS Lake Formation Tag-Based Access Control → | |:—|—:|