[Bug Fix] - gemini leaking FD for sync calls with litellm.completion

[ishaan-jaff]

[Bug Fix] - gemini leaking FD for sync calls with litellm.completion

Fixes this issue here: #8711 (comment)

Impact: Service crashes with "Too many open files" after processing multiple documents using Vertex AI/Gemini.

Root cause confirmed:

• Each call creates new HTTPHandler instances at lines 1485 and 1884 in vertex_and_google_ai_studio_gemini.py
• These HTTP clients are never closed, leaking file descriptors
• With parallel processing (10 workers × 20 LLM calls each), we quickly exhaust the FD limit

Relevant issues

Pre-Submission checklist

Please complete all items before asking a LiteLLM maintainer to review your PR

• I have Added testing in the tests/litellm/ directory, Adding at least 1 test is a hard requirement - see details
• I have added a screenshot of my new test passing locally
• My PR passes all unit tests on make test-unit
• My PR's scope is as isolated as possible, it only solves 1 specific problem

Type

🐛 Bug Fix
✅ Test

Changes

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
litellm	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Jul 21, 2025 8:59pm

[oneryalcin]

@ishaan-jaff I very much appreciate your speedy response 🙏 . This was awesome!