MLA-C01 Practice Question 79: Endpoint 4XX Errors
AWS Certified Machine Learning Engineer – Associate (MLA-C01) · Test 8 · question 9 of 10
Right after a new version of a mobile app is released, the SageMaker AI endpoint that powers its product recommendations shows a sharp rise in Invocation4XXErrors. Invocation5XXErrors stays at zero, ModelLatency and CPUUtilization are unchanged, and nothing about the endpoint, model, or container has been deployed or changed in a week. Older app versions still get recommendations normally.
What is the most likely cause?
Show answer and explanation
Answer:
Invocation4XXErrors counts InvokeEndpoint requests where the *model* returned a 4xx HTTP status, and 4xx codes mean the client sent a bad request. Invocation5XXErrors counts 5xx server-side failures. Since only 4XX errors rose, only requests from the new app version fail, and the endpoint itself has not changed, the evidence points to the new client. It is sending a payload or content type the container's input handler does not accept (option D).
- A is wrong because overload would show up in resource and latency metrics and typically as server-side errors or throttling. CPU and latency are unchanged, and older clients work fine. - B is wrong because an endpoint loads its model when instances start, not per request, and nothing has been redeployed. A serving problem would also hit every client and show up as 5XX errors. - C is wrong because a request denied by IAM is rejected by the SageMaker AI Runtime API before it reaches the model, so it would not appear in the model's Invocation4XXErrors metric at all.