Use executors in post-logging hooks (#14332)

litellm/caching/caching_handler.py

@@ -301,10 +301,12 @@ class LLMCachingHandler:
                         is_async=False,
                     )
 
-                    threading.Thread(
-                        target=logging_obj.success_handler,
-                        args=(cached_result, start_time, end_time, cache_hit),
-                    ).start()
+                    logging_obj.handle_sync_success_callbacks_for_async_calls(
+                        result=cached_result,
+                        start_time=start_time,
+                        end_time=end_time,
+                        cache_hit=cache_hit
+                    )
                     cache_key = litellm.cache._get_preset_cache_key_from_kwargs(
                         **kwargs
                     )
@@ -530,15 +532,17 @@ class LLMCachingHandler:
             end_time (datetime): The end time of the operation.
             cache_hit (bool): Whether it was a cache hit.
         """
-        asyncio.create_task(
-            logging_obj.async_success_handler(
-                cached_result, start_time, end_time, cache_hit
+        from litellm.litellm_core_utils.logging_worker import GLOBAL_LOGGING_WORKER
+
+        GLOBAL_LOGGING_WORKER.ensure_initialized_and_enqueue(
+            async_coroutine=logging_obj.async_success_handler(
+                result=cached_result, start_time=start_time, end_time=end_time, cache_hit=cache_hit
             )
         )
-        threading.Thread(
-            target=logging_obj.success_handler,
-            args=(cached_result, start_time, end_time, cache_hit),
-        ).start()
+
+        logging_obj.handle_sync_success_callbacks_for_async_calls(
+            result=cached_result, start_time=start_time, end_time=end_time, cache_hit=cache_hit
+        )
 
     async def _retrieve_from_cache(
         self, call_type: str, kwargs: Dict[str, Any], args: Tuple[Any, ...]

litellm/litellm_core_utils/litellm_logging.py

@@ -2774,6 +2774,7 @@ class Logging(LiteLLMLoggingBaseClass):
         result: Any,
         start_time: datetime.datetime,
         end_time: datetime.datetime,
+        cache_hit: Optional[Any] = None,
     ) -> None:
         """
         Handles calling success callbacks for Async calls.
@@ -2788,6 +2789,7 @@ class Logging(LiteLLMLoggingBaseClass):
             result,
             start_time,
             end_time,
+            cache_hit,
         )
 
     def _should_run_sync_callbacks_for_async_calls(self) -> bool: