From Legacy To Event Sourcing
summary
Practical account of moving a legacy Ruby/MySQL backend to event sourcing on the BEAM using classier for event streams and croisiere to unify hot Kafka and cold Postgres/S3 stores. Claims: per-partition ordering, upgrade blocks for schema migration (1.0 → 2.0), and a cache-backed GenServer-based user agents model to avoid slow replays. Worked solution: Confluent Kafka for ingestion (30‑day retention), then batch-move to Postgres and zipped file archives to satisfy GDPR and cost constraints.
AI-written summary · about
chapters (26)
- 0:00Intro
- 0:53Overview
- 1:32Free Event Sourcing
- 3:06Legacy vs Event Sourcing
- 3:27Event Sourcing Explained
- 6:17CQRS Explained
- 7:39DDD Explained
- 13:44Basic Event Sourcing
- 14:38Constraints
- 15:42Backend Support
- 17:01Practical Example
- 17:30Using Classier
- 18:43Adding a Field
- 20:02Default IP
- 23:09Upgrade Blocks
- 24:39Key Keys
- 26:02Partitioning
- 31:01Practical Limitations
- 34:49Where Are My Events
- 36:44How Do I Use It
- 37:39CQRS
- 38:18CQRS Agents
- 41:14Commands
- 41:32Agents
- 43:14Apply
- 44:58QA
