scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-10-08 03:08:45 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 116

Merge c524a885e3 into 49affa7c01

Browse source
This commit is contained in:
Yaroslav Ryzhan 2026-08-27 18:49:17 -05:00 • committed by GitHub
commit 832b175d04
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
2 changed files with 95 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
litellm/integrations/SlackAlerting/hanging_request_check.py
View file

@ -9,6 +9,8 @@ Notes:
import asyncio
import time
from collections.abc import Mapping
from types import MappingProxyType
from typing import TYPE_CHECKING, Any, Final
import litellm
@ -26,6 +28,8 @@ if TYPE_CHECKING:
else:
SlackAlerting = Any
_NO_METADATA: Mapping[str, Any] = MappingProxyType({})
class AlertingHangingRequestCheck:
"""
@ -57,7 +61,9 @@ class AlertingHangingRequestCheck:
if request_data is None:
return
request_metadata: Final = get_litellm_metadata_from_kwargs(kwargs=request_data)
request_metadata: Final = (
get_litellm_metadata_from_kwargs(kwargs=request_data) or request_data.get("metadata") or _NO_METADATA
)
model: Final = request_data.get("model", "")
api_base: str | None = None

88
tests/test_litellm/integrations/SlackAlerting/test_hanging_request_check.py
View file

@ -335,3 +335,91 @@ class TestAlertingHangingRequestCheck:
# Should not crash and should not send any alerts
hanging_request_checker.slack_alerting_object.send_alert.assert_not_called()
@pytest.mark.asyncio
async def test_add_request_uses_top_level_metadata_when_litellm_params_absent(
self, hanging_request_checker
):
"""Fall back to top-level metadata when litellm_params is absent."""
request_data = {
"litellm_call_id": "precall_request_001",
"model": "gpt-4",
"metadata": {
"user_api_key_alias": "my-key-alias",
"user_api_key_team_alias": "my-team-alias",
},
}
await hanging_request_checker.add_request_to_hanging_request_check(request_data)
cached_data = (
await hanging_request_checker.hanging_request_cache.async_get_cache(
key="precall_request_001"
)
)
assert cached_data is not None
assert cached_data.key_alias == "my-key-alias"
assert cached_data.team_alias == "my-team-alias"
@pytest.mark.asyncio
async def test_add_request_prefers_litellm_params_metadata_when_present(
self, hanging_request_checker
):
"""Prefer litellm_params.metadata over top-level metadata when present."""
request_data = {
"litellm_call_id": "postcall_request_002",
"model": "gpt-4",
"litellm_params": {
"metadata": {
"user_api_key_alias": "lp-key-alias",
"user_api_key_team_alias": "lp-team-alias",
},
},
"metadata": {
"user_api_key_alias": "should-not-use-this",
"user_api_key_team_alias": "should-not-use-this-either",
},
}
await hanging_request_checker.add_request_to_hanging_request_check(request_data)
cached_data = (
await hanging_request_checker.hanging_request_cache.async_get_cache(
key="postcall_request_002"
)
)
assert cached_data is not None
assert cached_data.key_alias == "lp-key-alias"
assert cached_data.team_alias == "lp-team-alias"
@pytest.mark.asyncio
async def test_add_request_preserves_team_alias_without_key_alias(
self, hanging_request_checker
):
"""Keep litellm_params team alias when key alias is absent (no overwrite)."""
request_data = {
"litellm_call_id": "team_only_request_003",
"model": "gpt-4",
"litellm_params": {
"metadata": {
"user_api_key_team_alias": "lp-team-alias",
},
},
"metadata": {
"user_api_key_team_alias": "should-not-use-this",
},
}
await hanging_request_checker.add_request_to_hanging_request_check(request_data)
cached_data = (
await hanging_request_checker.hanging_request_cache.async_get_cache(
key="team_only_request_003"
)
)
assert cached_data is not None
assert cached_data.key_alias == ""
assert cached_data.team_alias == "lp-team-alias"