add proctitle for tokenizers (#9952)

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

python/sglang/srt/entrypoints/http_server.py

@@ -29,6 +29,8 @@ import time
 from http import HTTPStatus
 from typing import Any, AsyncIterator, Callable, Dict, List, Optional
 
+import setproctitle
+
 # Fix a bug of Python threading
 setattr(threading, "_register_atexit", lambda *args, **kwargs: None)
 
@@ -1166,6 +1168,7 @@ def launch_server(
     2. Inter-process communication is done through IPC (each process uses a different port) via the ZMQ library.
     """
     if server_args.tokenizer_worker_num > 1:
+        setproctitle.setproctitle(f"sglang::http_server/multi_tokenizer_router")
         port_args = PortArgs.init_new(server_args)
         port_args.tokenizer_worker_ipc_name = (
             f"ipc://{tempfile.NamedTemporaryFile(delete=False).name}"
@@ -1174,6 +1177,7 @@ def launch_server(
             server_args=server_args, port_args=port_args
         )
     else:
+        setproctitle.setproctitle(f"sglang::http_server/tokenizer_manager")
         tokenizer_manager, template_manager, scheduler_info = _launch_subprocesses(
             server_args=server_args,
         )

python/sglang/srt/managers/multi_tokenizer_mixin.py

@@ -23,6 +23,7 @@ import threading
 from multiprocessing import shared_memory
 from typing import Dict
 
+import setproctitle
 import zmq
 import zmq.asyncio
 
@@ -476,6 +477,9 @@ class MultiTokenizerManager(TokenizerManager, MultiTokenizerMixin):
         server_args: ServerArgs,
         port_args: PortArgs,
     ):
+        setproctitle.setproctitle(
+            f"sglang::http_server/multi_tokenizer_manager:{os.getpid()}"
+        )
         # prevent init prefill bootstrapserver again
         disaggregation_mode = server_args.disaggregation_mode
         server_args.disaggregation_mode = "null"