[Feature] implement log channel separation and request log level system (#7190)

* feat: implement log channel separation and request log level system

* fix: log system improvements based on review

* add request_id to error logs, use RequestLogLevel enum, and unify logger implementation from utils to logger module

tests/scheduler/test_dp_scheduler.py

@@ -47,9 +47,13 @@ sys.modules["fastdeploy.engine.request"] = Mock()
 sys.modules["fastdeploy.scheduler"] = Mock()
 sys.modules["fastdeploy.scheduler.local_scheduler"] = Mock()
 sys.modules["fastdeploy.scheduler.data"] = Mock()
+sys.modules["fastdeploy.logger"] = Mock()
+sys.modules["fastdeploy.logger.request_logger"] = Mock()
 
 # Mock the get_logger function
 sys.modules["fastdeploy.utils"].get_logger = Mock(return_value=mock_logger)
+# Mock the log_request function
+sys.modules["fastdeploy.logger.request_logger"].log_request = Mock()
 
 
 # Mock the Request, RequestOutput, and ScheduledResponse classes
@@ -240,8 +244,9 @@ class TestDPLocalScheduler(unittest.TestCase):
 
     def test_put_results_with_finished_requests(self):
         """Test putting results with finished requests."""
-        # Reset mock logger
-        mock_logger.reset_mock()
+        # Get the mock log_request function
+        mock_log_request = sys.modules["fastdeploy.logger.request_logger"].log_request
+        mock_log_request.reset_mock()
 
         # Create mock request outputs
         results = [
@@ -254,13 +259,12 @@ class TestDPLocalScheduler(unittest.TestCase):
         with patch.object(self.scheduler, "responses_not_empty"):
             self.scheduler.put_results(results)
 
-        # Check that finished requests were logged - the logger should have been called
-        self.assertTrue(mock_logger.info.called)
+        # Check that finished requests were logged via log_request
+        self.assertTrue(mock_log_request.called)
         # Get the actual call arguments to verify the message format
-        call_args = mock_logger.info.call_args[0][0]
-        self.assertIn("finished responses", call_args)
-        self.assertIn("req1", call_args)
-        self.assertIn("req3", call_args)
+        call_kwargs = mock_log_request.call_args[1]
+        self.assertIn("finished responses", call_kwargs.get("message", ""))
+        self.assertIn("req1", str(call_kwargs.get("request_ids", [])))
 
     def test_put_results_with_new_responses(self):
         """Test putting results with new responses."""

tests/scheduler/test_local_scheduler.py

@@ -17,10 +17,11 @@ import unittest
 from unittest.mock import Mock, patch  # noqa: F401
 
 from fastdeploy.engine.request import Request, RequestMetrics, RequestOutput
+from fastdeploy.logger.request_logger import RequestLogLevel
 
 # Real FastDeploy imports
 from fastdeploy.scheduler.local_scheduler import LocalScheduler
-from fastdeploy.utils import envs, scheduler_logger
+from fastdeploy.utils import envs
 
 
 class TestLocalScheduler(unittest.TestCase):
@@ -221,9 +222,9 @@ class TestLocalScheduler(unittest.TestCase):
 
     def test_reset_logs_message(self):
         """Test that reset logs appropriate message."""
-        with patch.object(scheduler_logger, "info") as mock_info:
+        with patch("fastdeploy.scheduler.local_scheduler.log_request") as mock_log:
             self.scheduler.reset()
-            mock_info.assert_called_once_with("Scheduler has been reset")
+            mock_log.assert_called_once_with(RequestLogLevel.LIFECYCLE, message="Scheduler has been reset")
 
     def test_put_requests_duplicate_handling(self):
         """Test handling of duplicate request IDs."""
@@ -449,12 +450,14 @@ class TestLocalScheduler(unittest.TestCase):
         # Add request first
         self.scheduler.put_requests([self.mock_request_1])
 
-        with patch.object(scheduler_logger, "info") as mock_info:
+        with patch("fastdeploy.scheduler.local_scheduler.log_request") as mock_log_request:
             mock_output = self._create_test_request_output("req_1", finished=True)
             self.scheduler.put_results([mock_output])
 
-            # Should log finished response
-            self._assert_log_contains(mock_info, "finished responses")
+            # Should log finished response via log_request
+            mock_log_request.assert_called_once()
+            call_kwargs = mock_log_request.call_args[1]
+            self.assertIn("finished responses", call_kwargs.get("message", ""))
 
 
 if __name__ == "__main__":