An order service must notify inventory, shipping, and analytics independently when payment succeeds. The shipping service may be slow, but the order service should keep accepting new orders even if one consumer is unavailable. Which two changes best improve resilience? Select two.

Have the order service call all downstream services synchronously so failures are visible immediately.

Synchronous fan-out couples the order path to downstream availability and latency. If shipping or analytics is slow, the order request becomes slow or can fail even when the order itself was accepted successfully.

Use one shared SQS queue for all three consumers so they always process the same message.

A single shared queue creates a competing-consumers model, not fan-out. Only one consumer receives each message, so it does not independently deliver the same event to inventory, shipping, and analytics.

Store the event in a relational database and poll it from every consumer on a fixed schedule.

Polling a database adds latency, lock contention, and operational overhead. It also increases the chance that consumers fall behind or miss events during failures compared with a purpose-built messaging pattern.

What does this SAA-C03 question test?

Static NAT maps one inside address to one outside address.

What is the correct answer to this question?

The correct answer is: Publish the event to an Amazon SNS topic and subscribe a separate SQS queue for each downstream service. — SNS plus separate SQS queues gives each downstream service its own buffered work stream, so shipping being slow does not affect inventory or analytics. Adding a dead-letter queue per consumer isolates poison messages and prevents endless retries from consuming capacity. Together, these patterns improve both decoupling and operational resilience while keeping the order service responsive. Synchronous calls tightly couple the order path to downstream availability, which defeats the resilience goal. One shared queue makes consumers compete rather than independently receive events. Database polling adds delay and complexity, and it is not the best fit for event fan-out or failure isolation.

What should I do if I get this SAA-C03 question wrong?

Then try more questions from the same exam bank and focus on understanding why the wrong options are tempting.