The End of the Data Ice Age
summary
Argues that the hot-vs-cold data split is solvable by running per-entity state in memory and shows Java demos that implement event-sourced entities, key-value/value entities and views with Google protobuf APIs, plus actions (stateful serverless functions) wired via Kafka or Google Pub/Sub. Shows a working order-processing system on GitHub: millions of entity instances, events flowing between shopping-cart → order → shipping → stock, non-transactional stock allocation, and protobuf-generated stubs that remove explicit databases, brokers and Kubernetes plumbing.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:46Stateless pattern in backend apps
- 5:19Data on Ice concept
- 8:26Entity abstraction in code
- 11:38Event sourcing with entities
- 14:49Views and CQRS in abstraction
- 17:39Actions and event-driven systems
- 20:30Order processing demo
- 26:31Frictionless abstraction benefits
- 27:58End of the data Ice Age
- 29:22Freedom for event-driven systems
- 30:05Q&A
related talks
08mThe End of the Data Ice Age
25mAccelerating Data Processing and Consumption for Building Modern Apps at the Edge
49mTransforming Analytical Data with Event-Driven Architecture
55mThe Great Data Debate 2022
49mData Modeling in an Event Centric World
16mLife is But a (Data) Stream: Building Quality Data Pipelines
46mState or events? Which shall I keep?
39mThe Promise of Event Driven Programming with Fast Data