scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-13 23:11:40 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 101

test(e2e): trim the reliability helper docstrings to the cooldown rationale

Browse source
This commit is contained in:
mateo-berri 2026-09-11 19:32:40 -07:00
parent 9375719feb
commit a8ffc852f2
1 changed files with 3 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
tests/e2e/router/reliability_support.py
View file

@ -52,9 +52,8 @@ def create_bad_base_deployment(proxy: ProxyClient, name: str) -> str:
def create_never_benched_refusing_deployment(proxy: ProxyClient, name: str) -> str:
"""Register a deployment that refuses every call at the socket and opts out of the
stack's cooldown policy (cooldown_time 0), so the router keeps retrying it for the
whole run instead of benching it after allowed_fails and skipping the retry loop."""
"""cooldown_time 0 keeps the router retrying this deployment instead of benching it
after allowed_fails, which would skip the retry loop the memory test measures."""
return proxy.create_model(
name,
LiteLLMParamsBody(model=REAL_MODEL, api_key=REAL_KEY, api_base="http://127.0.0.1:9/v1", cooldown_time=0),
@ -124,8 +123,7 @@ def chat_override(
history: Sequence[ChatMessage] = (),
) -> StreamingResponse:
"""POST /chat/completions with an optional per-request router_settings_override,
returning the raw outcome so tests read status, body, and reliability headers.
`history` is the conversation sent ahead of the user turn carrying `content`."""
returning the raw outcome so tests read status, body, and reliability headers."""
return proxy.transport.send(
"/chat/completions",
headers=proxy.transport.bearer(key),