From d5080aeaa4d80f285d436ef66159fb2de4ffd3f7 Mon Sep 17 00:00:00 2001 From: Flora Feng <4florafeng@gmail.com> Date: Wed, 11 Mar 2026 03:11:41 -0400 Subject: [PATCH] [Refactor] Remove deadcode in Responses API serving (#36726) Signed-off-by: sfeng33 <4florafeng@gmail.com> Co-authored-by: Signed-off-by: yewentao256 --- vllm/entrypoints/openai/responses/serving.py | 23 -------------------- 1 file changed, 23 deletions(-) diff --git a/vllm/entrypoints/openai/responses/serving.py b/vllm/entrypoints/openai/responses/serving.py index a9356a8a4..ddd7bae04 100644 --- a/vllm/entrypoints/openai/responses/serving.py +++ b/vllm/entrypoints/openai/responses/serving.py @@ -1102,7 +1102,6 @@ class OpenAIServingResponses(OpenAIServing): event_deque: deque[StreamingResponsesResponse] = deque() new_event_signal = asyncio.Event() self.event_store[request.request_id] = (event_deque, new_event_signal) - response = None generator = self.responses_stream_generator(request, *args, **kwargs) try: async for event in generator: @@ -1111,15 +1110,6 @@ class OpenAIServingResponses(OpenAIServing): finally: new_event_signal.set() - if response is not None and isinstance(response, ErrorResponse): - # If the request has failed, update the status to "failed". - response_id = request.request_id - async with self.response_store_lock: - stored_response = self.response_store.get(response_id) - assert stored_response is not None - if stored_response.status not in ("completed", "cancelled"): - stored_response.status = "failed" - async def _run_background_request( self, request: ResponsesRequest, @@ -1226,19 +1216,6 @@ class OpenAIServingResponses(OpenAIServing): param="response_id", ) - def _make_store_not_supported_error(self) -> ErrorResponse: - return self.create_error_response( - err_type="invalid_request_error", - message=( - "`store=True` (default) is not supported. Please set " - "`store=False` in Responses API or set " - "`VLLM_ENABLE_RESPONSES_API_STORE=1` in the env var when " - "starting the vLLM server." - ), - status_code=HTTPStatus.BAD_REQUEST, - param="store", - ) - async def _process_simple_streaming_events( self, request: ResponsesRequest,