scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-10-07 02:59:05 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 103

fix(tests): fix flaky qwen global endpoint test by mocking AsyncHTTPHandler at class level

Browse source 
The test was creating a real AsyncHTTPHandler instance and patching its
post method, but the internal code creates its own handler, bypassing
the mock. This caused real API calls to Vertex AI, resulting in 401
auth errors in CI. Switched to patching AsyncHTTPHandler at the class
level, matching the pattern used by the passing GPT-OSS test.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
yuneng-jiang 2026-03-13 09:07:48 -07:00
parent b98ecd8c1d
commit 61cee53200
1 changed files with 15 additions and 22 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

37
tests/test_litellm/llms/vertex_ai/vertex_ai_partner_models/qwen/test_vertex_ai_qwen_global_endpoint.py
View file

@ -141,11 +141,6 @@ async def test_vertex_ai_qwen_global_endpoint_url():
"""
Test that Qwen models use the global endpoint URL.
"""
from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
from litellm.llms.vertex_ai.gemini.vertex_and_google_ai_studio_gemini import (
VertexLLM,
)
# Mock response
mock_response = MagicMock()
mock_response.status_code = 200
@ -168,35 +163,33 @@ async def test_vertex_ai_qwen_global_endpoint_url():
"usage": {"prompt_tokens": 10, "completion_tokens": 8, "total_tokens": 18},
}
client = AsyncHTTPHandler()
async def mock_post_func(*args, **kwargs):
return mock_response
mock_vertexai = MagicMock()
mock_vertexai.preview = MagicMock()
with patch.dict("sys.modules", {"vertexai": mock_vertexai}), patch.object(
client, "post", side_effect=mock_post_func
) as mock_post, patch.object(
VertexLLM, "_ensure_access_token", return_value=("fake-token", "test-project")
), patch.dict(
litellm.model_cost,
{"vertex_ai/qwen/qwen3-next-80b-a3b-instruct-maas": {"supported_regions": ["global"]}},
clear=False,
):
with patch("litellm.llms.custom_httpx.http_handler.AsyncHTTPHandler") as mock_http_handler, \
patch(
"litellm.llms.vertex_ai.gemini.vertex_and_google_ai_studio_gemini.VertexLLM._ensure_access_token",
return_value=("fake-token", "test-project"),
), \
patch.dict("sys.modules", {"vertexai": mock_vertexai, "vertexai.preview": mock_vertexai.preview}), \
patch.dict(
litellm.model_cost,
{"vertex_ai/qwen/qwen3-next-80b-a3b-instruct-maas": {"supported_regions": ["global"]}},
clear=False,
):
mock_http_handler.return_value.post = AsyncMock(return_value=mock_response)
response = await litellm.acompletion(
model="vertex_ai/qwen/qwen3-next-80b-a3b-instruct-maas",
messages=[{"role": "user", "content": "Hello"}],
vertex_ai_project="test-project",
client=client,
)
# Verify the mock was called
mock_post.assert_called_once()
mock_http_handler.return_value.post.assert_called_once()
# Get the call arguments
call_args = mock_post.call_args
call_args = mock_http_handler.return_value.post.call_args
called_url = call_args.kwargs["url"]
# Verify the URL uses global endpoint (no region prefix)