Loosely Coupled Lotteries & Cloudy Casinos: Nederlandse Loterij Case St.
summary
Custom integration platform running ~30 services from a monorepo handles spikes up to 2 million users (lottery draws, betting) and runs Java Spring Boot on AWS Kubernetes after migrating from ECS. Communication uses HTTP and Amazon SQS; Netflix OSS and Spring Cloud sleuth provide tracing. Gradle builds, tests and pushes Docker images in under 5 minutes via build caching. Central observability uses Datadog with JSON structured logging, MDC for request context, propagated trace IDs, and metrics for runtime scaling and debugging.
AI-written summary · about
chapters (21)
- 0:00Intro
- 3:29Project milestones
- 5:18Microservices: Common reasons
- 7:05Our reasons?
- 8:06Software architecture
- 9:37Microservices: Our reasons
- 15:59Resulting project setup
- 18:09Software architecture in practice
- 18:53System services (adapters)
- 23:13All-in-one services
- 24:41Resulting coupling
- 25:363 pillars of observability
- 26:15Patterns vs structured logging
- 27:10Logging appender configuration
- 28:03MDC
- 28:57Traces & slices
- 31:06Demo
- 34:30Metrics
- 35:50Metrics benefits
- 36:15Conclusion
- 37:50Outro
related talks
41mMicroservices in action at the DutchNationalPolice
01h 27mSpring Cloud at Netflix
47mApplying “web scale” patterns (...)
53mHow do we build high‑load microservices
40mCloud Native Java
01h 01mCloud Native Java
55mCloud Native Java with Josh Long
38mHighly Scalable Consumer Applications using Cloud Native Open Source Tools