How to migrate a mission critical railway system
summary
Migrates a mission‑critical NS execution system while keeping the old Oracle platform live by running the new stack in shadow production on Red Hat OpenShift across multiple data centers, using Spring Boot microservices, PostgreSQL, Red Hat AMQ and a proxy that forwards change messages plus DB context. A temporary adapter maps the complex legacy data model; a validator application compares outputs in real time (many messages per minute) and feeds a Kibana dashboard. Team of ~15 aims for multiple daily production deploys; supports 5,000 trains/day and ~700 train units.
AI-written summary · about
chapters (8)
- 0:00Introduction and speakers
- 1:51Mission critical railway systems
- 9:10System landscape overview
- 17:35Why rebuild from scratch
- 22:16Migration options explored
- 36:10Shadow production approach
- 44:28Project timeline and team
- 53:19Q&A and closing
related talks
32mChanging Engines Inflight - A Real Life Modernisation Story
24mMigrations under Production Load: How to Switch...
55mModernizing Legacy Code - While Not Bringing Down Prod
18mEngineering Resilience at Scale; Lessons from a Distributed Systems Transformation
28mApplication modernisation with Red Hat
41mInnovation takes courage
29mTaming the Hybrid Beast with DevOps
43mCRIMS: The case of the DB change that left no trace