xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-05-18 07:26:58 +08:00
Code Issues Packages Projects Releases Wiki Activity

[CI Bug] Fix Async Engine, Inputs, Utils, Worker Test: 'State' object has no attribute 'enable_server_load_tracking' (#20845)

Browse Source 
Signed-off-by: yewentao256 <zhyanwentao@126.com>
This commit is contained in:
Wentao Ye 2025-07-11 21:57:24 -04:00 committed by GitHub
parent 5f0af36af5
commit 7b828e30d5
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 12 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
vllm/entrypoints/utils.py
View File

@ -33,10 +33,12 @@ async def listen_for_disconnect(request: Request) -> None:
while True: while True:
message = await request.receive() message = await request.receive()
if message["type"] == "http.disconnect": if message["type"] == "http.disconnect":
if request.app.state.enable_server_load_tracking: # If load tracking is enabled *and* the counter exists, decrement
# on timeout/cancellation the BackgroundTask in load_aware_call # it. Combines the previous nested checks into a single condition
# cannot decrement the server load metrics. # to satisfy the linter rule.
# Must be decremented by with_cancellation instead. if (getattr(request.app.state, "enable_server_load_tracking",
False)
and hasattr(request.app.state, "server_load_metrics")):
request.app.state.server_load_metrics -= 1 request.app.state.server_load_metrics -= 1
break break
@ -101,9 +103,14 @@ def load_aware_call(func):
raise ValueError( raise ValueError(
"raw_request required when server load tracking is enabled") "raw_request required when server load tracking is enabled")
if not raw_request.app.state.enable_server_load_tracking: if not getattr(raw_request.app.state, "enable_server_load_tracking",
False):
return await func(*args, **kwargs) return await func(*args, **kwargs)
# ensure the counter exists
if not hasattr(raw_request.app.state, "server_load_metrics"):
raw_request.app.state.server_load_metrics = 0
raw_request.app.state.server_load_metrics += 1 raw_request.app.state.server_load_metrics += 1
try: try:
response = await func(*args, **kwargs) response = await func(*args, **kwargs)