fix: port already in use & nccl errors in gpu tests

src/lightning/fabric/CHANGELOG.md

@@ -27,7 +27,7 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ### Fixed
 
-- Fixed `EADDRINUSE` errors in distributed tests with port manager and retry logic ([#21309](https://github.com/Lightning-AI/pytorch-lightning/pull/21309))
+-
 
 
 ---

src/lightning/fabric/plugins/environments/lightning.py

@@ -13,11 +13,11 @@
 # limitations under the License.
 
 import os
+import socket
 
 from typing_extensions import override
 
 from lightning.fabric.plugins.environments.cluster_environment import ClusterEnvironment
-from lightning.fabric.utilities.port_manager import get_port_manager
 from lightning.fabric.utilities.rank_zero import rank_zero_only
 
 
@@ -104,38 +104,21 @@ def teardown(self) -> None:
         if "WORLD_SIZE" in os.environ:
             del os.environ["WORLD_SIZE"]
 
-        if self._main_port != -1:
-            get_port_manager().release_port(self._main_port)
-            self._main_port = -1
-
-        os.environ.pop("MASTER_PORT", None)
-        os.environ.pop("MASTER_ADDR", None)
-
 
 def find_free_network_port() -> int:
     """Finds a free port on localhost.
 
+    If the environment variable `STANDALONE_PORT` is set, its value is used as the port number.
+
     It is useful in single-node training when we don't want to connect to a real main node but have to set the
     `MASTER_PORT` environment variable.
 
-    The allocated port is reserved and won't be returned by subsequent calls until it's explicitly released.
-
-    Returns:
-        A port number that is reserved and free at the time of allocation
-
     """
-    # If an external launcher already specified a MASTER_PORT (for example, torch.distributed.spawn or
-    # multiprocessing helpers), reserve it through the port manager so no other test reuses the same number.
-    if "MASTER_PORT" in os.environ:
-        master_port_str = os.environ["MASTER_PORT"]
-        try:
-            existing_port = int(master_port_str)
-        except ValueError:
-            pass
-        else:
-            port_manager = get_port_manager()
-            if port_manager.reserve_existing_port(existing_port):
-                return existing_port
-
-    port_manager = get_port_manager()
-    return port_manager.allocate_port()
+    if "STANDALONE_PORT" in os.environ:
+        _port = os.environ["STANDALONE_PORT"]
+        return int(_port)
+    s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+    s.bind(("", 0))
+    port = s.getsockname()[1]
+    s.close()
+    return port