Log AI Gateway client HTTP errors

What does this MR do and why?

Previously:

  1. 4xx errors in the AI Gateway client would quietly be retried.
  2. 5xx errors would fail immediately without any indication.

This commit logs all AI Gateway HTTP errors with their status codes and bodies in llm.log to make it easier to troubleshoot server errors.

References

We are using this in Project Panda to troubleshoot upstream issues.

MR acceptance checklist

Please evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.

Screenshots or screen recordings

Screenshots are required for UI changes, and strongly recommended for all other merge requests.

Before After

How to set up and validate locally

  1. Set export AI_GATEWAY_URL=https://httpstat.us/500
  2. Invoke some AI Gateway request.
  3. View log/llm.log.
Edited by Stan Hu

Merge request reports

Loading