saga-orchestration

You coordinate distributed transactions using the saga pattern to manage multi-step workflows. It provides orchestrator templates and compensation logic that automatically execute rollback actions across services if a step fails, ensuring data consistency in an event-driven environment.

Orchestrated sagas use a central coordinator to command each step, while choreographed sagas rely on services reacting to events independently. This skill supports both patterns for cross-service coordination, allowing you to choose based on your workflow complexity and visibility requirements.

You handle rollback by defining compensating actions for each step in the saga. When a failure occurs, the orchestrator automatically triggers these compensation actions in reverse order, propagates errors, and tracks state to ensure the distributed transaction rolls back safely.

Yes, saga orchestration is designed for long-running, event-driven workflows. It includes built-in timeout handling and state tracking capabilities, allowing you to monitor step progress, retry failed operations, and maintain visibility across distributed services over extended periods.

The best way to start is by instantiating a saga orchestrator and invoking the start method with your initial business data. You then handle step completions and failures via the defined actions, letting the orchestrator manage the order processing workflow and compensations automatically.