event-driven

Event-driven architecture uses asynchronous messages to decouple distributed system components. You should use patterns like saga for distributed transactions and CQRS when separating read and write workloads in complex domains.

To design Kafka topics for exactly-once processing, you must implement specific consumer patterns and handle errors correctly. This Skill directs you to reference files detailing topic design and consumer error handling best practices.

The outbox pattern reliably publishes events within the same database transaction as your business logic. Reference documentation provides detailed guidance on implementing this pattern to ensure no events are lost during system failures.

Schema versioning manages structural changes to event messages without breaking existing consumers. The Skill's reference files provide specific guidance on schema versioning strategies to maintain compatibility across distributed system updates.

The guidance focuses on general event-driven patterns like outbox, saga, and CQRS alongside specific Kafka consumer patterns. Core distributed system design principles apply broadly, though consumer specifics target Kafka implementations.