Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .github/workflows/_base_test.yml
Original file line number Diff line number Diff line change
Expand Up @@ -272,7 +272,7 @@ jobs:

curl -X POST http://0.0.0.0:${FLASK_PORT}/switch \
-H "Content-Type: application/json" \
-d "{ \"--model\": \"/MODELDATA/ERNIE-4.5-0.3B-Paddle\", \"--max-concurrency\": 5000, \"--max-waiting-time\": 1 }"
-d "{ \"--model\": \"/MODELDATA/ERNIE-4.5-0.3B-Paddle\", \"--workers\": 1, \"--max-concurrency\": 5000, \"--max-waiting-time\": 1 }"
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

❓ 疑问 此处通过 --workers=1 修复了测试层的断言失败,但根本问题——多 worker 场景下 StatefulSemaphore 按 worker 数量等分信号量、导致请求分布不均时部分 worker 信号量提前耗尽——在生产环境中仍然存在。

建议方向:是否计划单独提 issue 跟踪生产侧的多 worker 信号量分配问题(如改为全局信号量或自适应分配策略)?

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

为什么会请求分配不均?

check_service 90
python -m pytest -sv test_max_waiting_time.py || TEST_EXIT_CODE=1

Expand Down
Loading