Threads not properly killed on windows

[racinmat]

🐛 Bug

On windows, killing

To Reproduce

I opened Lightning Studio but it redirected me to https://lightning.ai/templates. Below I include fully python code needed for replication.

Code sample

Code sample:

import os
import sys
import threading
import time

from litserve import LitAPI, LitServer

class _MinimalAPI(LitAPI):
    def setup(self, device: str) -> None:
        self.model = lambda x: x * 2

    def decode_request(self, request: dict) -> float:
        return float(request["input"])

    def predict(self, x: float) -> float:
        return self.model(x)

    def encode_response(self, output: float) -> dict:
        return {"output": output}

if __name__ == "__main__":
    print(f"[MAIN] PID={os.getpid()}  platform={sys.platform}")
    print("[MAIN] Starting LitServer...")
    server = LitServer(
        _MinimalAPI(),
        accelerator="cpu",
        devices=1,
        workers_per_device=1,
    )
    def _prompt():
        time.sleep(4)  # wait for uvicorn + inference workers to be ready
        print("[MAIN] Server fully running — click Stop in PyCharm to reproduce the bug.")

    threading.Thread(target=_prompt, daemon=True).start()
    server.run(port=8765, generate_client_file=False, log_level="info")

output when start debug mode and then click the Stop icon:

C:\Projects\others\LitServe\.venv\Scripts\python.exe -X pycache_prefix=C:\Users\Matej.Racinsky\AppData\Local\JetBrains\PyCharm2026.1\cpython-cache C:/Users/Matej.Racinsky/AppData/Local/Programs/PyCharm/plugins/python-ce/helpers/pydev/pydevd.py --multiprocess --qt-support=auto --client 127.0.0.1 --port 60226 --file C:\Projects\others\LitServe\litserve_win_shutdown2.py 
import sys; print('Python %s on %s' % (sys.version, sys.platform))
Connected to pydev debugger (build 261.24374.152)
[MAIN] PID=68424  platform=win32
[MAIN] Starting LitServer...
INFO:     Uvicorn running on http://0.0.0.0:8765 (Press CTRL+C to quit)
C:\Projects\others\LitServe\src\litserve\server.py:1500: UserWarning: Windows does not support forking. Using threads api_server_worker_type will be set to 'thread'
  warnings.warn(
[MAIN] Server fully running — click Stop in PyCharm NOW to reproduce / verify the fix.
Swagger UI is available at http://0.0.0.0:8765/docs
INFO:     Started server process [68424]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
2026-06-01 19:11:40,618 - MainProcess[68424] - litserve.server - INFO - server.py:1532 - KeyboardInterrupt received. Initiating graceful shutdown.
2026-06-01 19:11:40,618 - MainProcess[68424] - litserve.server - INFO - server.py:1190 - Shutting down LitServe...

the process keeps handing and never ends.

output when I run the code in PyChamr regularly and then click the Stop icon.

C:\Projects\others\LitServe\.venv\Scripts\python.exe C:\Projects\others\LitServe\litserve_win_shutdown2.py 
[MAIN] PID=34132  platform=win32
[MAIN] Starting LitServer...
INFO:     Uvicorn running on http://0.0.0.0:8765 (Press CTRL+C to quit)
C:\Projects\others\LitServe\src\litserve\server.py:1500: UserWarning: Windows does not support forking. Using threads api_server_worker_type will be set to 'thread'
  warnings.warn(
[MAIN] Server fully running — click Stop in PyCharm NOW to reproduce / verify the fix.
Swagger UI is available at http://0.0.0.0:8765/docs
INFO:     Started server process [34132]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
2026-06-01 19:17:06,069 - MainProcess[34132] - litserve.server - INFO - server.py:1532 - KeyboardInterrupt received. Initiating graceful shutdown.
2026-06-01 19:17:06,069 - MainProcess[34132] - litserve.server - INFO - server.py:1190 - Shutting down LitServe...
2026-06-01 19:17:06,069 - MainProcess[34132] - litserve.server - ERROR - server.py:1228 - Error during termination of LitServer-0 (PID: 57976): 'Thread' object has no attribute 'terminate'

The process ended, but I saw this error.

Expected behavior

I expect that in both cases, the process will end without any error.

Environment

If you published a Studio with your bug report, we can automatically get this information. Otherwise, please describe:

• PyTorch/Jax/Tensorflow Version (e.g., 1.0): None needed
• OS (e.g., Linux): Windows 11
• How you installed PyTorch (conda, pip, source): not used
• Build command you used (if compiling from source): uv sync
• Python version: 3.10.11
• CUDA/cuDNN version: none
• GPU models and configuration: none
• Any other relevant information: none

Additional context

Full environment: I used main branch of LitServe and reproduced it, but I also replicated it with version 0.2.17.

Here is my pip freeze:

aiofile==3.9.0
aiohappyeyeballs==2.6.2
aiohttp==3.13.5
aiosignal==1.4.0
annotated-doc==0.0.4
annotated-types==0.7.0
anyio==4.13.0
asgi-lifespan==2.1.0
async-timeout==5.0.1
attrs==26.1.0
Authlib==1.7.2
backports.asyncio.runner==1.2.0
backports.tarfile==1.2.0
beartype==0.22.9
cachetools==7.1.4
caio==0.9.25
certifi==2026.5.20
cffi==2.0.0
charset-normalizer==3.4.7
click==8.4.1
colorama==0.4.6
coverage==7.14.1
cryptography==48.0.0
cyclopts==4.16.1
distro==1.9.0
dnspython==2.8.0
docstring_parser==0.18.0
email-validator==2.3.0
exceptiongroup==1.3.1
fastapi==0.136.3
fastmcp==3.3.1
fastmcp-slim==3.3.1
filelock==3.29.0
frozenlist==1.8.0
fsspec==2026.4.0
griffelib==2.0.2
h11==0.16.0
hf-xet==1.5.0
httpcore==1.0.9
httptools==0.8.0
httpx==0.28.1
httpx-sse==0.4.3
huggingface_hub==1.17.0
idna==3.17
importlib_metadata==9.0.0
iniconfig==2.3.0
jaraco.classes==3.4.0
jaraco.context==6.1.2
jaraco.functools==4.5.0
Jinja2==3.1.6
jiter==0.15.0
joserfc==1.6.8
jsonargparse==4.49.0
jsonref==1.1.0
jsonschema==4.26.0
jsonschema-path==0.5.0
jsonschema-specifications==2025.9.1
keyring==25.7.0
librt==0.11.0
lightning==2.6.5
lightning-utilities==0.15.3
-e git+ssh://git@github.com/Lightning-AI/LitServe.git@a69b6354f634f636298daba2637d68075bf5b681#egg=litserve
markdown-it-py==4.2.0
MarkupSafe==3.0.3
mcp==1.27.2
mdurl==0.1.2
more-itertools==11.1.0
mpmath==1.3.0
multidict==6.7.1
mypy==1.20.0
mypy_extensions==1.1.0
networkx==3.4.2
numpy==2.2.6
openai==2.38.0
openapi-pydantic==0.5.1
opentelemetry-api==1.42.1
packaging==26.2
pathable==0.6.0
pathspec==1.1.1
pillow==12.2.0
platformdirs==4.10.0
pluggy==1.6.0
propcache==0.5.2
psutil==7.2.2
py-key-value-aio==0.4.5
pycparser==3.0
pydantic==2.13.4
pydantic-settings==2.14.1
pydantic_core==2.46.4
Pygments==2.20.0
PyJWT==2.13.0
pyperclip==1.11.0
pytest==9.0.3
pytest-asyncio==1.4.0
pytest-cov==7.1.0
pytest-retry==1.7.0
python-dotenv==1.2.2
python-multipart==0.0.30
pytorch-lightning==2.6.5
pywin32==311
pywin32-ctypes==0.2.3
PyYAML==6.0.3
pyzmq==27.1.0
referencing==0.37.0
regex==2026.5.9
requests==2.34.2
rich==15.0.0
rich-rst==2.0.1
rpds-py==0.30.0
safetensors==0.7.0
shellingham==1.5.4
sniffio==1.3.1
sse-starlette==3.4.4
starlette==1.2.1
sympy==1.14.0
tenacity==9.1.4
tokenizers==0.22.2
tomli==2.4.1
torch==2.12.0
torchmetrics==1.9.0
torchvision==0.27.0
tqdm==4.67.3
transformers==5.9.0
typer==0.25.1
typing-inspection==0.4.2
typing_extensions==4.15.0
uncalled-for==0.3.2
urllib3==2.7.0
uvicorn==0.48.0
watchfiles==1.2.0
websockets==16.0
yarl==1.24.2
zipp==4.1.0