[Bugfix] Fix some cases of The client socket has timed out after 600s while trying to connect to

vllm/envs.py

@@ -4,6 +4,7 @@
 
 if TYPE_CHECKING:
     VLLM_HOST_IP: str = ""
+    HOST_IP: str = ""
     VLLM_PORT: Optional[int] = None
     VLLM_RPC_BASE_PATH: str = tempfile.gettempdir()
     VLLM_USE_MODELSCOPE: bool = False
@@ -153,7 +154,13 @@ def get_default_config_root():
     # If you are using multi-node inference, you should set this differently
     # on each node.
     'VLLM_HOST_IP':
-    lambda: os.getenv('VLLM_HOST_IP', "") or os.getenv("HOST_IP", ""),
+    lambda: os.getenv('VLLM_HOST_IP', ""),
+
+    # used in distributed environment to determine the ip address
+    # of the current node, when the env variable VLLM_HOST_IP is not set
+    # and the it's hard to get the host ip.
+    'HOST_IP':
+    lambda: os.getenv('HOST_IP', ""),
 
     # used in distributed environment to manually set the communication port
     # Note: if VLLM_PORT is set, and some code asks for multiple ports, the

vllm/utils.py

@@ -479,6 +479,11 @@ def get_ip() -> str:
     except Exception:
         pass
 
+    try:
+        return socket.gethostbyname(socket.gethostname())
+    except Exception:
+        pass
+
     # try ipv6
     try:
         s = socket.socket(socket.AF_INET6, socket.SOCK_DGRAM)
@@ -489,12 +494,15 @@ def get_ip() -> str:
     except Exception:
         pass
 
-    warnings.warn(
-        "Failed to get the IP address, using 0.0.0.0 by default."
-        "The value can be set by the environment variable"
-        " VLLM_HOST_IP or HOST_IP.",
-        stacklevel=2)
-    return "0.0.0.0"
+    if envs.HOST_IP:
+        return envs.HOST_IP
+    else:
+        warnings.warn(
+            "Failed to get the IP address, using 0.0.0.0 by default."
+            "The value can be set by the environment variable"
+            " VLLM_HOST_IP or HOST_IP.",
+            stacklevel=2)
+        return "0.0.0.0"
 
 
 def is_valid_ipv6_address(address: str) -> bool: