xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-10 08:04:58 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bugfix] Bind api server port before starting engine (#8491)

Browse Source
This commit is contained in:
Kevin Lin 2024-09-16 15:56:28 -05:00 committed by GitHub
parent 2759a43a26
commit 47f5e03b5b
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 6 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
vllm/entrypoints/openai/api_server.py
View File

@ -5,6 +5,7 @@ import multiprocessing
import os
import re
import signal
import socket
import tempfile
from argparse import Namespace
from contextlib import asynccontextmanager
@ -525,6 +526,9 @@ async def run_server(args, **uvicorn_kwargs) -> None:
logger.info("vLLM API server version %s", VLLM_VERSION)
logger.info("args: %s", args)
temp_socket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
temp_socket.bind(("", args.port))
def signal_handler(*_) -> None:
# Interrupt server on sigterm while initializing
raise KeyboardInterrupt("terminated")
@ -541,6 +545,8 @@ async def run_server(args, **uvicorn_kwargs) -> None:
model_config = await async_engine_client.get_model_config()
init_app_state(async_engine_client, model_config, app.state, args)
temp_socket.close()
shutdown_task = await serve_http(
app,
limit_concurrency=async_engine_client.limit_concurrency,