Add comprehensive test coverage for engine_client.py to improve code …

.github/workflows/_unit_test_coverage.yml

@@ -41,7 +41,7 @@ jobs:
 
   run_tests_with_coverage:
     runs-on: [self-hosted, GPU-h1z1-2Cards]
-    timeout-minutes: 90
+    timeout-minutes: 120
     needs: check_cov_skip
     if: needs.check_cov_skip.outputs.can-skip != 'true'
     outputs:

tests/entrypoints/test_engine_client.py

@@ -143,6 +143,17 @@ def mock_fd_config_with_eplb():
 class TestEngineClient(unittest.IsolatedAsyncioTestCase):
     async def asyncSetUp(self):
         """Set up test fixtures before each test method."""
+        # Mock main_process_metrics to avoid KeyError
+        self.metrics_mock = Mock()
+        self.metrics_mock.request_params_max_tokens = Mock()
+        self.metrics_mock.prompt_tokens_total = Mock()
+        self.metrics_mock.request_prompt_tokens = Mock()
+        self.metrics_patcher = patch(
+            "fastdeploy.entrypoints.engine_client.main_process_metrics",
+            self.metrics_mock,
+        )
+        self.metrics_patcher.start()
+
         # Create a properly configured tokenizer mock first
         mock_tokenizer = Mock()
         mock_tokenizer.sp_model = Mock()
@@ -285,6 +296,11 @@ def test_engine_client_initialized_by_fd_config(self):
         self.engine_client.clear_update_lock.__enter__ = Mock(return_value=None)
         self.engine_client.clear_update_lock.__exit__ = Mock(return_value=None)
 
+    async def asyncTearDown(self):
+        """Clean up test fixtures after each test method."""
+        if hasattr(self, "metrics_patcher"):
+            self.metrics_patcher.stop()
+
     async def test_add_request(self):
         request = {
             "request_id": "test-request-id",
@@ -309,6 +325,17 @@ class TestEngineClientValidParameters(unittest.TestCase):
 
     def setUp(self):
         """Set up test fixtures for valid_parameters tests"""
+        # Mock main_process_metrics to avoid KeyError
+        self.metrics_mock = Mock()
+        self.metrics_mock.request_params_max_tokens = Mock()
+        self.metrics_mock.prompt_tokens_total = Mock()
+        self.metrics_mock.request_prompt_tokens = Mock()
+        self.metrics_patcher = patch(
+            "fastdeploy.entrypoints.engine_client.main_process_metrics",
+            self.metrics_mock,
+        )
+        self.metrics_patcher.start()
+
         # Mock the dependencies
         mock_tokenizer = MagicMock()
         mock_tokenizer.sp_model = MagicMock()
@@ -396,6 +423,11 @@ def setUp(self):
                             self.engine_client.prefix_tree_status_signal = Mock()
                             self.engine_client.prefix_tree_status_signal.value = np.array([0])
 
+    def tearDown(self):
+        """Clean up test fixtures after each test method."""
+        if hasattr(self, "metrics_patcher"):
+            self.metrics_patcher.stop()
+
     def test_max_logprobs_valid_values(self):
         """Test valid max_logprobs values"""
         # Test positive max_logprobs