diff --git a/litellm/router_utils/pre_call_checks/encrypted_content_affinity_check.py b/litellm/router_utils/pre_call_checks/encrypted_content_affinity_check.py index 39f72d01341..806c3db7601 100644 --- a/litellm/router_utils/pre_call_checks/encrypted_content_affinity_check.py +++ b/litellm/router_utils/pre_call_checks/encrypted_content_affinity_check.py @@ -385,11 +385,20 @@ class EncryptedContentAffinityCheck(CustomLogger): model_id, ) - excluded_ids: Final = frozenset( - as_excluded_deployment_ids(typed_request_kwargs.get("_excluded_deployment_ids")) + previous_response_id: Final = typed_request_kwargs.get("previous_response_id") + previous_response_origin: Final = ResponsesAPIRequestUtils.get_model_id_from_response_id( + None if previous_response_id is None else str(previous_response_id) ) - retry_skipped_ids: Final = frozenset( - as_retry_skipped_deployment_ids(typed_request_kwargs.get("_retry_skipped_deployment_ids")) + previous_response_binds_origin: Final = previous_response_origin == model_id + excluded_ids: Final[frozenset[str]] = ( + frozenset() + if previous_response_binds_origin + else frozenset(as_excluded_deployment_ids(typed_request_kwargs.get("_excluded_deployment_ids"))) + ) + retry_skipped_ids: Final[frozenset[str]] = ( + frozenset() + if previous_response_binds_origin + else frozenset(as_retry_skipped_deployment_ids(typed_request_kwargs.get("_retry_skipped_deployment_ids"))) ) not_excluded_candidates: Final = [ candidate for candidate in candidates if self._deployment_model_id(candidate) not in excluded_ids diff --git a/tests/unit/router_utils/pre_call_checks/test_encrypted_content_affinity_check.py b/tests/unit/router_utils/pre_call_checks/test_encrypted_content_affinity_check.py index 4c804ccdc09..a3745c7b23f 100644 --- a/tests/unit/router_utils/pre_call_checks/test_encrypted_content_affinity_check.py +++ b/tests/unit/router_utils/pre_call_checks/test_encrypted_content_affinity_check.py @@ -2633,6 +2633,10 @@ def _reasoning_input_minted_by(origin: str) -> list[dict[str, object]]: ] +def _previous_response_id_from(origin: str) -> str: + return ResponsesAPIRequestUtils._build_responses_api_response_id("openai", origin, "resp_test") + + _ORIGIN_FAILED_THIS_REQUEST: Final = pytest.mark.parametrize( "failed_in_request", [{"_excluded_deployment_ids": ["region-a"]}, {"_retry_skipped_deployment_ids": ("region-a",)}], @@ -2641,8 +2645,9 @@ _ORIGIN_FAILED_THIS_REQUEST: Final = pytest.mark.parametrize( @pytest.mark.asyncio @_ORIGIN_FAILED_THIS_REQUEST +@pytest.mark.parametrize("continuing", [{}, {"previous_response_id": _previous_response_id_from("region-b")}]) async def test_affinity_strips_and_dispatches_when_origin_already_failed_this_request( - failed_in_request: dict[str, object], + failed_in_request: dict[str, object], continuing: dict[str, object] ): from litellm.router_utils.pre_call_checks.encrypted_content_affinity_check import ( EncryptedContentAffinityCheck, @@ -2656,13 +2661,40 @@ async def test_affinity_strips_and_dispatches_when_origin_already_failed_this_re model="gpt-5.4", healthy_deployments=group, messages=None, - request_kwargs={"litellm_metadata": {}, "input": reasoning_input, **failed_in_request}, + request_kwargs={"litellm_metadata": {}, "input": reasoning_input, **failed_in_request, **continuing}, ) assert result is group assert reasoning_input[0] == {"type": "reasoning", "summary": [{"type": "summary_text", "text": "scattering"}]} +@pytest.mark.asyncio +@_ORIGIN_FAILED_THIS_REQUEST +async def test_affinity_keeps_origin_and_reasoning_when_previous_response_id_binds_the_failed_origin( + failed_in_request: dict[str, object], +): + from litellm.router_utils.pre_call_checks.encrypted_content_affinity_check import ( + EncryptedContentAffinityCheck, + ) + + pool: Final = (_region_deployment("region-a"), _region_deployment("region-b")) + reasoning_input: Final = _reasoning_input_minted_by("region-a") + original_reasoning: Final = dict(reasoning_input[0]) + + assert await EncryptedContentAffinityCheck(router=litellm.Router(model_list=list(pool))).async_filter_deployments( + model="gpt-5.4", + healthy_deployments=list(pool), + messages=None, + request_kwargs={ + "litellm_metadata": {}, + "input": reasoning_input, + "previous_response_id": _previous_response_id_from("region-a"), + **failed_in_request, + }, + ) == [pool[0]] + assert reasoning_input[0] == original_reasoning + + @pytest.mark.asyncio @pytest.mark.parametrize("regions", [("region-a",), ("region-a", "region-b")]) async def test_affinity_keeps_reasoning_when_every_deployment_already_refused_the_request(regions: tuple[str, ...]):