Question 630
For the enterprise requirement "Payment message consumers can fail repeatedly on one malformed message and must isolate it without blocking all valid messages", the team implemented this change: Configure SQS redrive policy with an appropriate dead-letter queue and maxReceiveCount. What should operations verify instead of relying on configuration alone?
Explanation
Correct Answer
This measures the behavior required by the specific solution, rather than simply checking that it was configured. SQS dead-letter queues preserve failed messages for separate inspection and remediation. The acceptance condition is linked directly to the scenario: Payment message consumers can fail repeatedly on one malformed message and must isolate it without blocking all valid messages.
Incorrect Answers
Events captured during the fault can be replayed to a corrected consumer after recovery. However, this does not establish the required outcome: a poison message goes to the DLQ after configured attempts while valid following messages proceed.
A simulated transient error is retried according to policy and a persistent error follows a defined failure branch. However, this does not establish the required outcome: a poison message goes to the DLQ after configured attempts while valid following messages proceed.
The API meets latency SLOs under bursts and incurs low idle compute spend without permanently allocated hosts. However, this does not establish the required outcome: a poison message goes to the DLQ after configured attempts while valid following messages proceed.