Legacy Architecture Migration Patterns with DDD
summary
Migration decisions hinge on concrete trade-offs: PayFit uses domain events, change data capture, AWS Step Functions, Snowflake, Datadog and structured logging plus an AI debug agent to trace cross‑system failures; a concrete claim is that up to 80% features unused makes selective migration essential. A façade translates legacy events into a new collaborator model, which unlocks other domains but creates unexpected three‑way sync pain, higher maintenance and duplicate monthly events. Migrating writes first for collaborators succeeds; contracts use a different, safer pattern.
AI-written summary · about
chapters (10)
- 0:00Legacy migration patterns
- 1:31Migration principles
- 7:32Migration patterns trade-offs
- 12:46Strangler fig and bubble patterns
- 20:09Payit case study
- 25:45Domain model drift
- 32:28Observability in migrations
- 37:43Model synchronization
- 40:53Legacy event publishing
- 43:16Q&A
related talks
58mJourney to the Cloud - a Rearchitecting Story
14mFrom Brownfield to Greenfield
25mData-Driven Migration from Monoliths to Event-Driven Microservices
36mLegacy code - lessons learned during migration (...)
23mLegacy to Cloud: Transforming Insurance Systems with .NET Microservices & Azure
48mArchitecture Modernization: Aligning Software, Strategy, and Structure
01h 06mRewriting Legacy Code
28mDaniel Coulbourne :: Flip the Switch Slowly