implement-event-sourcing

You implement event sourcing by creating an AggregateRoot that validates commands and emits events, then publishing those events to a reactive EventStore over R2DBC. Fast reads are enabled by replaying events and building projections from snapshots.

The AggregateRoot pattern validates business invariants inside public command methods. Upon successful validation, it emits domain events via applyChange, ensuring state transitions are captured immutably for reliable audit trails and temporal queries.

Yes, you can use snapshots and multi-tenancy in a reactive Java service. Snapshots optimize aggregate loading by periodically saving state, while TenantContext isolates tenant data through Reactor context propagation in Mono and Flux streams.

Event upcasting transforms older event versions into newer schemas during event replay. This allows your event-sourced application to evolve domain models over time without breaking existing EventStore records or requiring data migrations.

Yes, the Firefly Framework supports a transactional outbox for event-sourced microservices. It ensures reliable persistence by atomically writing events to the EventStore alongside business state changes, preventing data loss across distributed services.