fix(messages): name the deployment on a rust stream's transport error (#51358)

When the upstream connection drops or stalls mid-stream on the Rust Messages
path, the caller gets the raw httpx error without litellm_debug_info, so the
llm_exceptions alert names no API base, model group or deployment. Python's
stream iterator attaches that info to the same error, and the Rust host now does
too.

litellm/rust_bridge/host/routes/messages.py

@@ -28,6 +28,7 @@ from litellm.litellm_core_utils.core_helpers import (
     proxy_request_headers,
     stream_cost_injection_requested,
 )
+from litellm.litellm_core_utils.exception_mapping_utils import with_deployment_debug_info
 from litellm.litellm_core_utils.get_llm_provider_logic import get_provider_affinity_header
 from litellm.litellm_core_utils.get_provider_specific_headers import ProviderSpecificHeaderUtils
 from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
@@ -768,6 +769,20 @@ def _log_resolved_call(request: Mapping[str, object], request_provider: str) ->
     )
 
 
+def _with_deployment_debug_info(error: Exception, request: Mapping[str, object], request_provider: str) -> Exception:
+    logger: Final = request.get("litellm_logging_obj")
+    if not isinstance(logger, LiteLLMLoggingObj):
+        return error
+    details: Final = logger.model_call_details
+    provider: Final = details.get("custom_llm_provider")
+    return with_deployment_debug_info(
+        error,
+        logger.model,
+        provider if isinstance(provider, str) and provider else request_provider,
+        details.get("litellm_params"),
+    )
+
+
 def map_failure(
     error: Exception, request: Mapping[str, object], request_provider: str, asynchronous: bool = False
 ) -> Exception:
@@ -775,12 +790,16 @@ def map_failure(
     if getattr(error, "messages_timeout", None) == "Body":
         # Python reads a stream's body outside its HTTP client's error mapping, so the transport's
         # own timeout reaches the caller unmapped.
-        return httpx.ReadTimeout(_BODY_READ_TIMEOUT if http_transport.uses_aiohttp_transport() else "")
+        return _with_deployment_debug_info(
+            httpx.ReadTimeout(_BODY_READ_TIMEOUT if http_transport.uses_aiohttp_transport() else ""),
+            request,
+            request_provider,
+        )
     body_cut: Final = getattr(error, "messages_body_cut", None)
     if body_cut is not None and request.get("stream") is True:
         # Python reads a stream's body outside its HTTP client's error mapping, so a body the
         # connection cut short reaches the caller as its transport raised it.
-        return _body_cut(*_BODY_CUT.validate_python(body_cut))
+        return _with_deployment_debug_info(_body_cut(*_BODY_CUT.validate_python(body_cut)), request, request_provider)
     transport_timeout: Final = _TRANSPORT_TIMEOUTS.get(getattr(error, "messages_timeout", None))
     if transport_timeout is not None:
         # Python's `litellm.Timeout` is raised while handling httpx's own timeout, which the proxy's

tests/unit/rust_bridge/host/routes/test_messages.py

@@ -523,6 +523,41 @@ def test_a_complete_body_cut_is_a_synthesized_server_error(
     assert text in str(mapped)
 
 
+def _deployment_logger() -> Logging:
+    logger: Final = Logging(
+        model=_ADAPTIVE_OPUS,
+        messages=[],
+        stream=True,
+        call_type="anthropic_messages",
+        start_time=datetime.datetime.now(),
+        litellm_call_id="cut",
+        function_id="cut",
+    )
+    logger.model_call_details["custom_llm_provider"] = "anthropic"
+    logger.model_call_details["litellm_params"] = {
+        "api_base": "https://upstream.example/v1",
+        "metadata": {"model_group": "claude-group", "deployment": "anthropic/claude-opus-4-6"},
+    }
+    return logger
+
+
+@pytest.mark.parametrize("marker", ("body_cut", "body_timeout"))
+def test_a_stream_transport_error_names_the_deployment_like_the_python_stream_iterator(marker: str) -> None:
+    native: Final = RuntimeError("upstream network error")
+    if marker == "body_cut":
+        native.messages_body_cut = (19, None, False)  # pyright: ignore[reportAttributeAccessIssue]  # marker the native host sets
+    else:
+        native.messages_timeout = "Body"  # pyright: ignore[reportAttributeAccessIssue]  # marker the native host sets
+    request: Final = {"model": _ADAPTIVE_OPUS, "stream": True, "litellm_logging_obj": _deployment_logger()}
+
+    debug_info: Final = getattr(messages_host.map_failure(native, request, "anthropic"), "litellm_debug_info", None)
+
+    assert isinstance(debug_info, str)
+    assert "API Base: `https://upstream.example/v1`" in debug_info
+    assert "model_group: `claude-group`" in debug_info
+    assert "deployment: `anthropic/claude-opus-4-6`" in debug_info
+
+
 _TRANSPORT: Final = "error sending request: client error"