diff --git a/tests/unit/proxy/pass_through_endpoints/test_pass_through_endpoints.py b/tests/unit/proxy/pass_through_endpoints/test_pass_through_endpoints.py index 0f7bbf156db..3ce34c69b1c 100644 --- a/tests/unit/proxy/pass_through_endpoints/test_pass_through_endpoints.py +++ b/tests/unit/proxy/pass_through_endpoints/test_pass_through_endpoints.py @@ -30,6 +30,7 @@ from starlette.datastructures import UploadFile as StarletteUploadFile import litellm from litellm._logging import verbose_proxy_logger +from litellm.caching.dual_cache import DualCache from litellm.constants import DEFAULT_REQUEST_TIMEOUT_SECONDS from litellm.integrations.custom_guardrail import CustomGuardrail from litellm.integrations.custom_logger import CustomLogger @@ -65,6 +66,7 @@ from litellm.proxy.pass_through_endpoints.success_handler import ( PassThroughEndpointLogging, ) from litellm.proxy.route_llm_request import ProxyModelNotFoundError +from litellm.proxy.utils import ProxyLogging from litellm.types import utils as types_utils from litellm.types.guardrails import GuardrailEventHooks from litellm.types.passthrough_endpoints.pass_through_endpoints import ( @@ -442,7 +444,7 @@ async def test_pass_through_request_failure_handler(): assert failure["user_api_key_dict"] is user_api_key_dict assert failure["original_exception"] is request_failure - assert "traceback_str" in failure + assert "Request failed" in failure["traceback_str"] assert failure["request_data"]["contents"] == "not a list" assert failure["request_data"]["model"] == "" @@ -8528,6 +8530,24 @@ _RELAYED_VERTEX_COUNT_TOKENS_TARGET = ( ) +class _FailureHookRecorder(CustomLogger): + def __init__(self): + super().__init__() + self.failures = [] + + async def async_post_call_failure_hook( + self, request_data, original_exception, user_api_key_dict, traceback_str=None + ): + self.failures.append( + { + "request_data": request_data, + "original_exception": original_exception, + "user_api_key_dict": user_api_key_dict, + "traceback_str": traceback_str, + } + ) + + async def _run_passthrough_failure( target: str, custom_llm_provider: str | None, @@ -8535,21 +8555,22 @@ async def _run_passthrough_failure( stream: bool = False, user_api_key_dict: UserAPIKeyAuth | None = None, ) -> dict: + recorder = _FailureHookRecorder() with ( - pytest.MonkeyPatch.context() as transport_patch, - patch("litellm.proxy.proxy_server.proxy_logging_obj") as mock_proxy_logging, + pytest.MonkeyPatch.context() as litellm_patch, + patch( # test-quality-ok: proxy_logging_obj is a proxy_server module global read inside pass_through_request; there is no injection seam + "litellm.proxy.proxy_server.proxy_logging_obj", ProxyLogging(user_api_key_cache=DualCache()) + ), respx.mock(assert_all_called=False) as upstream_mock, ): - transport_patch.setattr(litellm, "disable_aiohttp_transport", True) + litellm_patch.setattr(litellm, "disable_aiohttp_transport", True) + litellm_patch.setattr(litellm, "callbacks", [recorder]) litellm.in_memory_llm_clients_cache.flush_cache() upstream_route = upstream_mock.route(method="POST") if isinstance(upstream, Exception): upstream_route.mock(side_effect=upstream) else: upstream_route.mock(return_value=upstream) - mock_proxy_logging.pre_call_hook = AsyncMock(side_effect=lambda **kwargs: kwargs["data"]) - mock_proxy_logging.post_call_failure_hook = AsyncMock() - mock_proxy_logging.post_call_response_headers_hook = AsyncMock(return_value=None) mock_request = MagicMock(spec=Request) mock_request.method = "POST" proxy_url = httpx.URL(target.replace("https://", "http://test-proxy.com/")) @@ -8574,8 +8595,8 @@ async def _run_passthrough_failure( pass await asyncio.sleep(0) litellm.in_memory_llm_clients_cache.flush_cache() - mock_proxy_logging.post_call_failure_hook.assert_called_once() - return mock_proxy_logging.post_call_failure_hook.call_args.kwargs + assert len(recorder.failures) == 1 + return recorder.failures[0] def _upstream_400(target: str) -> httpx.Response: