AI-103: Developing AI Apps and Agents on Azure

Question

You have a Microsoft Foundry project that contains an agent. The agent uses threads and file uploads, and it calls an Azure OpenAI model deployment.

During load testing, calls intermittently fail and return an HTTP 429 rate limit exceeded error. Some user uploads fail and generate an HTTP 400 file size exceeded error.

You need to mitigate the errors and reduce call failures. The solution must remain within the service and model limits.

What should you do to resolve each error? Drag the appropriate actions to the correct errors. Each action may be used once, more than once, or not at all.

Drag & Drop

HTTP 429:

HTTP 400:

Explanation

Microsoft Foundry Agent Service documentation explicitly defines a limit error reference table: a 429 rate_limit_exceeded error from a model deployment is mitigated by implementing exponential backoff with jitter in the client's retry logic, which spaces out retries to let throttling recover without exceeding the tokens-per-minute/requests-per-minute quota. A 400 file_size_exceeded error from file uploads is mitigated by splitting the content into smaller files (or reducing file size) before uploading, since the per-file size limit is a fixed service limit that cannot be increased. Both remedies work within the existing service and model limits, as required, rather than requesting new quota (which is scoped to the subscription/deployment, not tenant-wide, and is a separate action from reducing failures within current limits).

Learn more

AI-103: Developing AI Apps and Agents on Azure

AstroTutor

Notes

Readiness

QuestionQ94

Community Discussion