test: fix v2 transport intermittent test failure (#29002)

Only relying on the number of peers for detecting a new connection
suffers from race conditions, as unrelated previous peers could
disconnect at anytime in-between. Use the more robust approach of
watching for an increased highest peer id instead (again using the
`getpeerinfo` RPC call), with a newly introduced context manager
method `TestNode.wait_for_new_peer()`.

Fixes #29009.

test/functional/test_framework/test_node.py

@@ -519,6 +519,24 @@ class TestNode():
             'Expected messages "{}" does not partially match log:\n\n{}\n\n'.format(
                 str(expected_msgs), print_log))
 
+    @contextlib.contextmanager
+    def wait_for_new_peer(self, timeout=5):
+        """
+        Wait until the node is connected to at least one new peer. We detect this
+        by watching for an increased highest peer id, using the `getpeerinfo` RPC call.
+        Note that the simpler approach of only accounting for the number of peers
+        suffers from race conditions, as disconnects from unrelated previous peers
+        could happen anytime in-between.
+        """
+        def get_highest_peer_id():
+            peer_info = self.getpeerinfo()
+            return peer_info[-1]["id"] if peer_info else -1
+
+        initial_peer_id = get_highest_peer_id()
+        yield
+        wait_until_helper_internal(lambda: get_highest_peer_id() > initial_peer_id,
+                                   timeout=timeout, timeout_factor=self.timeout_factor)
+
     @contextlib.contextmanager
     def profile_with_perf(self, profile_name: str):
         """