scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-11-14 09:03:16 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 214

Merge 34e1b75222 into 4a7fb30b0a

Browse source
This commit is contained in:
Shelton 2026-11-04 12:40:51 -08:00 • committed by GitHub
commit 9e229ff712
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
2 changed files with 20 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
litellm/responses/streaming_iterator.py
View file

@ -250,6 +250,8 @@ _ERROR_CODE_HTTP_STATUS: Final[Mapping[str, int]] = MappingProxyType(
"server_error": 500,
"rate_limit_exceeded": 429,
"insufficient_quota": 429,
"resource_unavailable": 429,
"flex_unavailable": 429,
"vector_store_timeout": 504,
"invalid_prompt": 400,
"invalid_image": 400,

18
tests/unit/responses/test_streaming_iterator_error_events.py
View file

@ -75,6 +75,24 @@ def _make_error_chunk(error_type: str, code: str, message: str = "err") -> Error
return ErrorEvent(type=ResponsesAPIStreamEvents.ERROR, sequence_number=0, error=error_obj)
@pytest.mark.parametrize(
"error_type,error_code",
(
("resource_unavailable", "flex_unavailable"),
("resource_unavailable", None),
(None, "flex_unavailable"),
),
)
def test_streamed_flex_unavailable_maps_to_429(error_type: str | None, error_code: str | None) -> None:
# OpenAI returns 429 for flex-capacity exhaustion on non-streamed requests; the streamed
# /v1/responses error event must map to 429 as well, not fall through to 500 (#49736).
assert _status_code_for_error_fields(error_type, error_code) == 429
def test_streamed_unrelated_server_error_still_maps_to_500() -> None:
assert _status_code_for_error_fields("server_error", "server_error") == 500
@pytest.mark.parametrize("provider", ("openai", "azure", "bedrock"))
@pytest.mark.parametrize("header", ("x-amzn-RequestId", "x-request-id", None))
@pytest.mark.parametrize(