CDC Analytics With Graph Stream Zoomer
summary
A distributed system named Graph Stream Zoomer groups Neo4j CDC into time windows and emits aggregated graph summaries; it reads enriched transaction logs via the Kafka connector, runs grouping and aggregate functions on Apache Flink, and pushes results to a TCP socket for a yFiles map. Demo uses New York bike trips: 30‑second windows produce grouped edges with counts and min/max/avg durations, supports Neo4j CDC diff/full modes and pattern filters, is horizontally scalable, open source on GitHub, and the demo visualization updates correctly in real time.
AI-written summary · about
chapters (12)
- 0:00Introduction and presenters
- 1:57Graph stream basics
- 4:26Change data capture explained
- 7:23Enabling CDC in Neo4j
- 9:07CDC functions and examples
- 11:39Kafka CDC connector setup
- 14:09Grouping graph streams
- 18:21Craft Stream Zoomer framework
- 21:10CDC analytics use cases
- 24:01Demo architecture overview
- 25:05Live demo walkthrough
- 27:58Closing and acknowledgments
related talks
21mSerious Network Analysis Using Hadoop & Neo4j
25mBuilding an Entire Graph Application with Nodestream
35mBuilding Real-Time Graph Analytics Platforms
30mChange Data Streaming Patterns in Distributed Systems
49mLet’s Build a Streaming Data Pipeline!
49mModern Graph Databases
47mElasticSearch, MongoDB and Neo4j walk into a bar a tale of different databases
41mChange Data Streaming Patterns in Distributed Systems