update startup.py:

1. controller改为自动等待
2. 修复：分部分启动时进程引用错误

startup.py

@@ -210,7 +210,7 @@ def _set_app_seq(app: FastAPI, q: Queue, run_seq: int):
             q.put(run_seq)
 
 
-def run_controller(q: Queue, run_seq: int = 1, log_level: str = "INFO"):
+def run_controller(q: Queue, run_seq: int = 1, log_level: str = "INFO", e: mp.Event = None):
     import uvicorn
     import httpx
     from fastapi import Body
@@ -223,6 +223,11 @@ def run_controller(q: Queue, run_seq: int = 1, log_level: str = "INFO"):
     )
     _set_app_seq(app, q, run_seq)
 
+    @app.on_event("startup")
+    def on_startup():
+        if e is not None:
+            e.set()
+
     # add interface to release and load model worker
     @app.post("/release_worker")
     def release_worker(
@@ -517,8 +522,9 @@ async def start_main_server():
     signal.signal(signal.SIGTERM, handler("SIGTERM"))
 
     mp.set_start_method("spawn")
+    manager = mp.Manager()
 
-    queue = Queue()
+    queue = manager.Queue()
     args, parser = parse_args()
 
     if args.all_webui:
@@ -558,11 +564,12 @@ async def start_main_server():
     else:
         log_level = "INFO"
 
+    controller_started = manager.Event()
     if args.openai_api:
         process = Process(
             target=run_controller,
-            name=f"controller({os.getpid()})",
-            args=(queue, process_count() + 1, log_level),
+            name=f"controller",
+            args=(queue, process_count() + 1, log_level, controller_started),
             daemon=True,
         )
 
@@ -570,7 +577,7 @@ async def start_main_server():
 
         process = Process(
             target=run_openai_api,
-            name=f"openai_api({os.getpid()})",
+            name=f"openai_api",
             args=(queue, process_count() + 1),
             daemon=True,
         )
@@ -581,7 +588,7 @@ async def start_main_server():
         if not config.get("online_api"):
             process = Process(
                 target=run_model_worker,
-                name=f"model_worker - {args.model_name} ({os.getpid()})",
+                name=f"model_worker - {args.model_name}",
                 args=(args.model_name, args.controller_address, queue, process_count() + 1, log_level),
                 daemon=True,
             )
@@ -594,7 +601,7 @@ async def start_main_server():
             if config.get("online_api") and config.get("worker_class"):
                 process = Process(
                     target=run_model_worker,
-                    name=f"model_worker - {model_name} ({os.getpid()})",
+                    name=f"model_worker - {model_name}",
                     args=(model_name, args.controller_address, queue, process_count() + 1, log_level),
                     daemon=True,
                 )
@@ -604,7 +611,7 @@ async def start_main_server():
     if args.api:
         process = Process(
             target=run_api_server,
-            name=f"API Server{os.getpid()})",
+            name=f"API Server",
             args=(queue, process_count() + 1),
             daemon=True,
         )
@@ -614,7 +621,7 @@ async def start_main_server():
     if args.webui:
         process = Process(
             target=run_webui,
-            name=f"WEBUI Server{os.getpid()})",
+            name=f"WEBUI Server",
             args=(queue, process_count() + 1),
             daemon=True,
         )
@@ -626,17 +633,30 @@ async def start_main_server():
     else:
         try:
             # 保证任务收到SIGINT后，能够正常退出
-            processes["controller"].start()
-            # 使用await asyncio.sleep(3)可以让后续代码等待一段时间
-            await asyncio.sleep(3)
-
-            processes["openai_api"].start()
-            processes["model_worker"].start()
-            for p in processes["online-api"]:
+            if p:= processes.get("controller"):
                 p.start()
+                p.name = f"{p.name} ({p.pid})"
+                controller_started.wait()
 
-            processes["api"].start()
-            processes["webui"].start()
+            if p:= processes.get("openai_api"):
+                p.start()
+                p.name = f"{p.name} ({p.pid})"
+
+            if p:= processes.get("model_worker"):
+                p.start()
+                p.name = f"{p.name} ({p.pid})"
+
+            for p in processes.get("online-api", []):
+                p.start()
+                p.name = f"{p.name} ({p.pid})"
+
+            if p:= processes.get("api"):
+                p.start()
+                p.name = f"{p.name} ({p.pid})"
+
+            if p:= processes.get("webui"):
+                p.start()
+                p.name = f"{p.name} ({p.pid})"
 
             while True:
                 no = queue.get()
@@ -647,17 +667,18 @@ async def start_main_server():
                 else:
                     queue.put(no)
 
-            if model_worker_process := processes.pop("model_worker", None):
+            if model_worker_process := processes.get("model_worker"):
                 model_worker_process.join()
-            for process in processes.pop("online-api", []):
+            for process in processes.get("online-api", []):
                 process.join()
             for name, process in processes.items():
+                if name not in ["model_worker", "online-api"]:
                     if isinstance(p, list):
                         for work_process in p:
                             work_process.join()
                     else:
                         process.join()
-        except:
+        except Exception as e:
             # if model_worker_process := processes.pop("model_worker", None):
             #     model_worker_process.terminate()
             # for process in processes.pop("online-api", []):
@@ -669,6 +690,7 @@ async def start_main_server():
             #             work_process.terminate()
             #     else:
             #         process.terminate()
+            logger.error(e)
             logger.warning("Caught KeyboardInterrupt! Setting stop event...")
         finally:
             # Send SIGINT if process doesn't exit quickly enough, and kill it as last resort
@@ -703,6 +725,8 @@ if __name__ == "__main__":
         asyncio.set_event_loop(loop)
     # 同步调用协程代码
     loop.run_until_complete(start_main_server())
+
+
 # 服务启动后接口调用示例：
 # import openai
 # openai.api_key = "EMPTY" # Not support yet