fix(proxy/utils.py): log rejected proxy requests to langfuse

litellm/integrations/langfuse.py

@@ -84,6 +84,7 @@ class LangFuseLogger:
             print_verbose(
                 f"Langfuse Logging - Enters logging function for model {kwargs}"
             )
+
             litellm_params = kwargs.get("litellm_params", {})
             metadata = (
                 litellm_params.get("metadata", {}) or {}
@@ -373,7 +374,11 @@ class LangFuseLogger:
                 # just log `litellm-{call_type}` as the generation name
                 generation_name = f"litellm-{kwargs.get('call_type', 'completion')}"
 
+            if response_obj is not None and "system_fingerprint" in response_obj:
                 system_fingerprint = response_obj.get("system_fingerprint", None)
+            else:
+                system_fingerprint = None
+
             if system_fingerprint is not None:
                 optional_params["system_fingerprint"] = system_fingerprint
 

litellm/proxy/proxy_server.py

@@ -3663,6 +3663,17 @@ async def chat_completion(
         if data["model"] in litellm.model_alias_map:
             data["model"] = litellm.model_alias_map[data["model"]]
 
+        ## LOGGING OBJECT ## - initialize logging object for logging success/failure events for call
+        data["litellm_call_id"] = str(uuid.uuid4())
+        logging_obj, data = litellm.utils.function_setup(
+            original_function="acompletion",
+            rules_obj=litellm.utils.Rules(),
+            start_time=litellm.utils.get_utc_datetime(),
+            **data,
+        )
+
+        data["litellm_logging_obj"] = logging_obj
+
         ### CALL HOOKS ### - modify incoming data before calling the model
         data = await proxy_logging_obj.pre_call_hook(
             user_api_key_dict=user_api_key_dict, data=data, call_type="completion"

litellm/proxy/utils.py

@@ -1,6 +1,6 @@
 from typing import Optional, List, Any, Literal, Union
 import os, subprocess, hashlib, importlib, asyncio, copy, json, aiohttp, httpx, time
-import litellm, backoff
+import litellm, backoff, traceback
 from litellm.proxy._types import (
     UserAPIKeyAuth,
     DynamoDBArgs,
@@ -199,6 +199,33 @@ class ProxyLogging:
             print_verbose(f"final data being sent to {call_type} call: {data}")
             return data
         except Exception as e:
+            if "litellm_logging_obj" in data:
+                logging_obj: litellm.utils.Logging = data["litellm_logging_obj"]
+
+                ## ASYNC FAILURE HANDLER ##
+                error_message = ""
+                if isinstance(e, HTTPException):
+                    if isinstance(e.detail, str):
+                        error_message = e.detail
+                    elif isinstance(e.detail, dict):
+                        error_message = json.dumps(e.detail)
+                    else:
+                        error_message = str(e)
+                else:
+                    error_message = str(e)
+                error_raised = Exception(f"{error_message}")
+                await logging_obj.async_failure_handler(
+                    exception=error_raised,
+                    traceback_exception=traceback.format_exc(),
+                )
+
+                ## SYNC FAILURE HANDLER ##
+                try:
+                    logging_obj.failure_handler(
+                        error_raised, traceback.format_exc()
+                    )  # DO NOT MAKE THREADED - router retry fallback relies on this!
+                except Exception as error_val:
+                    pass
             raise e
 
     async def during_call_hook(

litellm/tests/test_function_setup.py

@@ -25,7 +25,7 @@ def test_empty_content():
         pass
 
     function_setup(
-        original_function=completion,
+        original_function="completion",
         rules_obj=rules_obj,
         start_time=datetime.now(),
         messages=[],

litellm/utils.py

@@ -2433,7 +2433,7 @@ class Rules:
 ####### CLIENT ###################
 # make it easy to log if completion/embedding runs succeeded or failed + see what happened | Non-Blocking
 def function_setup(
-    original_function, rules_obj, start_time, *args, **kwargs
+    original_function: str, rules_obj, start_time, *args, **kwargs
 ):  # just run once to check if user wants to send their data anywhere - PostHog/Sentry/Slack/etc.
     try:
         global callback_list, add_breadcrumb, user_logger_fn, Logging
@@ -2457,10 +2457,12 @@ def function_setup(
             len(litellm.input_callback) > 0
             or len(litellm.success_callback) > 0
             or len(litellm.failure_callback) > 0
-        ) and len(callback_list) == 0:
+        ) and len(
+            callback_list  # type: ignore
+        ) == 0:  # type: ignore
             callback_list = list(
                 set(
-                    litellm.input_callback
+                    litellm.input_callback  # type: ignore
                     + litellm.success_callback
                     + litellm.failure_callback
                 )
@@ -2469,7 +2471,7 @@ def function_setup(
         ## ASYNC CALLBACKS
         if len(litellm.input_callback) > 0:
             removed_async_items = []
-            for index, callback in enumerate(litellm.input_callback):
+            for index, callback in enumerate(litellm.input_callback):  # type: ignore
                 if inspect.iscoroutinefunction(callback):
                     litellm._async_input_callback.append(callback)
                     removed_async_items.append(index)
@@ -2480,7 +2482,7 @@ def function_setup(
 
         if len(litellm.success_callback) > 0:
             removed_async_items = []
-            for index, callback in enumerate(litellm.success_callback):
+            for index, callback in enumerate(litellm.success_callback):  # type: ignore
                 if inspect.iscoroutinefunction(callback):
                     litellm._async_success_callback.append(callback)
                     removed_async_items.append(index)
@@ -2496,7 +2498,7 @@ def function_setup(
 
         if len(litellm.failure_callback) > 0:
             removed_async_items = []
-            for index, callback in enumerate(litellm.failure_callback):
+            for index, callback in enumerate(litellm.failure_callback):  # type: ignore
                 if inspect.iscoroutinefunction(callback):
                     litellm._async_failure_callback.append(callback)
                     removed_async_items.append(index)
@@ -2539,7 +2541,7 @@ def function_setup(
             user_logger_fn = kwargs["logger_fn"]
         # INIT LOGGER - for user-specified integrations
         model = args[0] if len(args) > 0 else kwargs.get("model", None)
-        call_type = original_function.__name__
+        call_type = original_function
         if (
             call_type == CallTypes.completion.value
             or call_type == CallTypes.acompletion.value
@@ -2721,7 +2723,7 @@ def client(original_function):
         try:
             if logging_obj is None:
                 logging_obj, kwargs = function_setup(
-                    original_function, rules_obj, start_time, *args, **kwargs
+                    original_function.__name__, rules_obj, start_time, *args, **kwargs
                 )
             kwargs["litellm_logging_obj"] = logging_obj
 
@@ -3030,7 +3032,7 @@ def client(original_function):
         try:
             if logging_obj is None:
                 logging_obj, kwargs = function_setup(
-                    original_function, rules_obj, start_time, *args, **kwargs
+                    original_function.__name__, rules_obj, start_time, *args, **kwargs
                 )
             kwargs["litellm_logging_obj"] = logging_obj