Stream Processing at the Edge with Apache Kafka
summary
Shows a working edge pipeline: an ESP32 sensor posting JSON over HTTP to an HTTP bridge, Kafka brokers run on a three-node Raspberry Pi k3s cluster with Strimzi, a Quarkus app uses the Kafka Streams API to compute one-minute windowed averages, MirrorMaker replicates aggregated topics to a central cluster, and Argo/Flux-style GitOps plus operators handle deployment and certs. Demo uses microPython on the device, k3s for Kubernetes, and real hardware, proving edge processing reduces latency and cloud egress while surviving intermittent connectivity.
AI-written summary · about
chapters (14)
- 0:00Stream processing intro
- 2:32Stream processing in industries
- 5:43Edge data processing
- 7:30Edge environments for Kafka
- 8:51Kafka for edge processing
- 12:02Edge stream processing architecture
- 15:50Kubernetes operators for edge
- 18:02Demo setup overview
- 21:57Edge Kubernetes cluster
- 23:21Raw sensor data in Kafka
- 24:04Kafka Streams API
- 25:30Centralized edge visualization
- 27:18Demo recap
- 27:39Q&A
related talks
39mAdvanced Stream Processing on the Edge
03h 00mStream Processing, Going Beyond Pub/Sub with Kafka Streams and...
25mAccelerating Data Processing and Consumption for Building Modern Apps at the Edge
25mKubernetes IoT Edge Working Group: Using Event Driven Architecture at Edge
24mStreaming analytics on edge with Kuiper and KubeEdge
01h 29mEvent stream processing using Kafka Streams
01h 00mIn Depth Kafka Streams
35mBuilding Smart Farms with Edge Computing Systems Using Kubernetes
