From 3146519add735bc51a6a983af9e9c4a8b8d3373e Mon Sep 17 00:00:00 2001 From: Nick Hill Date: Fri, 1 Aug 2025 15:37:55 +0100 Subject: [PATCH] [BugFix] Don't change title of top-level process (#22032) Signed-off-by: Nick Hill --- vllm/entrypoints/cli/serve.py | 11 ++++++----- vllm/entrypoints/openai/api_server.py | 4 ++-- 2 files changed, 8 insertions(+), 7 deletions(-) diff --git a/vllm/entrypoints/cli/serve.py b/vllm/entrypoints/cli/serve.py index 0305354a66e85..9762a1de9edd3 100644 --- a/vllm/entrypoints/cli/serve.py +++ b/vllm/entrypoints/cli/serve.py @@ -18,7 +18,8 @@ from vllm.entrypoints.utils import (VLLM_SUBCMD_PARSER_EPILOG, show_filtered_argument_or_group_from_help) from vllm.logger import init_logger from vllm.usage.usage_lib import UsageContext -from vllm.utils import FlexibleArgumentParser, decorate_logs, get_tcp_uri +from vllm.utils import (FlexibleArgumentParser, decorate_logs, get_tcp_uri, + set_process_title) from vllm.v1.engine.core import EngineCoreProc from vllm.v1.engine.utils import CoreEngineProcManager, launch_core_engines from vllm.v1.executor.abstract import Executor @@ -74,7 +75,7 @@ def run_headless(args: argparse.Namespace): if args.api_server_count > 1: raise ValueError("api_server_count can't be set in headless mode") - # set_process_title("Headless_ProcManager") + # Create the EngineConfig. engine_args = vllm.AsyncEngineArgs.from_cli_args(args) usage_context = UsageContext.OPENAI_API_SERVER @@ -139,8 +140,6 @@ def run_multi_api_server(args: argparse.Namespace): orig_disable_mm_preprocessor_cache = args.disable_mm_preprocessor_cache - # set_process_title("ProcManager") - if num_api_servers > 1: setup_multiprocess_prometheus() @@ -225,7 +224,9 @@ def run_api_server_worker_proc(listen_address, **uvicorn_kwargs) -> None: """Entrypoint for individual API server worker processes.""" - # Add process-specific prefix to stdout and stderr. + # Set process title and add process-specific prefix to stdout and stderr. + server_index = client_config.get("client_index", 0) if client_config else 0 + set_process_title("APIServer", str(server_index)) decorate_logs() uvloop.run( diff --git a/vllm/entrypoints/openai/api_server.py b/vllm/entrypoints/openai/api_server.py index 26db1357da4d0..1be03c57a1f1b 100644 --- a/vllm/entrypoints/openai/api_server.py +++ b/vllm/entrypoints/openai/api_server.py @@ -102,7 +102,7 @@ from vllm.transformers_utils.tokenizer import MistralTokenizer from vllm.usage.usage_lib import UsageContext from vllm.utils import (Device, FlexibleArgumentParser, decorate_logs, get_open_zmq_ipc_path, is_valid_ipv6_address, - set_process_title, set_ulimit) + set_ulimit) from vllm.v1.metrics.prometheus import get_prometheus_registry from vllm.version import __version__ as VLLM_VERSION @@ -1824,7 +1824,7 @@ async def run_server_worker(listen_address, ToolParserManager.import_tool_parser(args.tool_parser_plugin) server_index = client_config.get("client_index", 0) if client_config else 0 - set_process_title("APIServer", str(server_index)) + # Load logging config for uvicorn if specified log_config = load_log_config(args.log_config_file) if log_config is not None: