Event-Driven Architecture

Event-Driven Architecture: A Production Handbook is a practical, in-depth guide to building distributed systems that communicate through events - the architecture behind modern platforms at scale. Written for engineers who have to run these systems in production, not just design them on a whiteboard, it moves past theory into the decisions and trade-offs that determine whether an event-driven system stays reliable under real load. Across 20 chapters, the book covers the full lifecycle of an event-driven system: choosing between Kafka, Pulsar, and NATS JetStream; designing partitioning and ordering guarantees; tuning producers for durability and throughput; managing consumer groups and rebalancing; and reasoning clearly about delivery semantics - at-most-once, at-least-once, and exactly-once - and what each actually costs. It goes deep on schema evolution and registries, event sourcing and CQRS, sagas for distributed transactions, and the transactional outbox pattern for reliably publishing events from a database. Later chapters cover the operational reality often left out of architecture books: idempotency and deduplication, retries and dead-letter queues, retention and compaction, observability, security, and testing strategies including contract testing. Every chapter includes working code examples, failure-mode callouts, and production checklists - a field manual for engineers building, running, and debugging event-driven systems at scale, not just another survey of concepts.