분산 트랜잭션을 위한 사가 패턴

In a monolith one database transaction can update everything atomically. With microservices each service owns its own database, so a single ACID transaction across them is not possible.

A saga is a sequence of local transactions, one per service. Each step publishes an event that triggers the next. If a step fails, earlier steps are undone with compensating actions.

An order saga might be: reserve inventory, charge payment, schedule shipping. Each is a local transaction in its own service, linked together as one logical workflow.

There is no rollback across services, so you compensate. If payment fails after inventory was reserved, you run a compensating action to release the inventory.

In choreography, services react to each other events with no central coordinator. The inventory service emits InventoryReserved; the payment service listens and reacts.

Choreography is decoupled and simple for short flows, but the overall workflow is implicit and hard to follow as steps grow. Debugging means tracing events across many services.

In orchestration, a central orchestrator tells each service what to do and tracks progress. The workflow lives in one place, making it easier to reason about.

Events may be delivered more than once. Every step and compensation must be idempotent so retries do not double-charge or double-reserve. Use a unique operation id to dedupe.

The orchestrator persists the current step so it can resume after a crash. Store saga state durably and mark each step as pending, completed, or compensated.

If a compensation itself fails, retry it with backoff and alert operators. Sagas guarantee eventual consistency, not instant consistency, so design the UI to reflect in-progress states.

Use sagas when a business process spans multiple services and must remain consistent. For a single service, a normal database transaction is simpler and preferred. Do not add saga complexity unless you actually cross service boundaries.

Check your saga knowledge.

You learned the saga pattern for cross-service consistency:

• Break a transaction into local steps with compensating actions
• Coordinate via choreography (events) or orchestration (central coordinator)
• Make steps idempotent and persist saga state
• Accept eventual consistency

Sagas keep distributed workflows reliable without a global transaction.