apache / apache/tooling-llmao

A backend error arrives as HTTP 200 with a null body

Open
#32 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
1
Forks
1
PR merge metrics
No merged PRs in 30d

Description

vLLM returned 422 (it could not fetch an image URL -- Wikimedia 403s server-side fetches). Through the proxy that became 200 with a body of literal null.

A caller gets no status, no error object, nothing to branch on. It crashed the benchmark harness, which is the mild version; a client library would likely return None and fail later somewhere unrelated.

Expected: the upstream status and error body pass through.

Contributor guide

No contributing guide indexed for this repository

Research direction

Start at the proxy's request and upstream-error handling entry point, then reproduce the vLLM response with a 422 status and error body. Done means the proxy returns that upstream status and body instead of HTTP 200 with a literal null; add or run the relevant request test if one exists.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
api, backend
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Active
Clarity
Mostly clear
Newbie friendliness
65/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.