REST APIs vs. Data Streams: Which Is Right for Your Business?
summary
Argues that replacing webs of point‑to‑point REST calls with Kafka dramatically cuts integration complexity — a 4‑system bidirectional mesh balloons into nearly 40,000 integrations at scale — and shows concrete tooling: Kafka Connect for ingestion, stream processing for joins and time windows, and Connect with Confluent / stream sharing for external SaaS links. Recommends CDC from legacy DBs to Kafka, claims gigabytes‑per‑second, millions of transactions per second throughput, and shows the pattern working end‑to‑end.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:06REST APIs vs data streams matter
- 3:54REST APIs in integration challenges
- 5:43Monoliths to microservices
- 7:50Cost of point-to-point integrations
- 9:58REST APIs' hidden complexity
- 10:39Sync vs async integration
- 12:03What is Kafka
- 13:50Kafka uses TCP
- 14:55Stream processors vs REST APIs
- 15:38Decoupling with Kafka
- 16:42REST APIs and Kafka coexist
- 17:45Stream sharing as data product
- 19:10Integration choices revisited
- 19:52Modern integration takeaways
related talks
06mJay Kreps interview
23mREST the events: REST APIs for Event-Driven Architecture
50mApplication development and data in the emerging world ...
39mData Streaming? I don't even know her
47mStream All the Things!
27mNavigating Streaming Infrastructure
26mLeveraging Data in Motion in a Cloud-first World
45mApplications in the Emerging World of Stream Processing