resonate-saga-pattern-typescript

You orchestrate distributed transactions by defining a TypeScript-based saga with ordered forward steps and corresponding compensations, running it through a context that automatically unwinds progress in reverse order upon failure.

The saga pattern coordinates long-running distributed transactions by breaking work into steps with compensating actions, ensuring eventual consistency when failures occur across multi-service workflows like order processing, payments, and fulfillment.

Cross-service rollback is handled by executing compensations in reverse order, while idempotency is supported natively so that retrying failed steps or timeouts does not duplicate side effects during distributed transaction recovery.

The best way to manage timeouts and retry policies in a TypeScript saga is using an orchestration-first approach with centralized control, explicit error handling, and configurable retry policies for each step.

Yes, you can persist saga state for observability, as the orchestration approach supports optional persistence to track multi-step workflow progress, forward steps, and compensating actions across distributed services.

No, you do not need a separate event broker, as this orchestration-first approach uses centralized control to coordinate saga steps, compensating actions, and error handling directly within your TypeScript application.