xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-10 10:30:37 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fix raw_request extraction in load_aware_call decorator (#15382)

Browse Source 
Signed-off-by: Daniel Salib <danielsalib@meta.com>
This commit is contained in:
daniel-salib 2025-03-25 22:29:54 -07:00 committed by GitHub
parent 6c663dfd5e
commit 5aefd6ac31
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 10 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
vllm/entrypoints/utils.py
View File

@ -68,13 +68,20 @@ def decrement_server_load(request: Request):
def load_aware_call(func):
@functools.wraps(func)
async def wrapper(*args, raw_request: Request, **kwargs):
async def wrapper(*args, **kwargs):
raw_request = kwargs.get("raw_request",
args[1] if len(args) > 1 else None)
if raw_request is None:
raise ValueError(
"raw_request required when server load tracking is enabled")
if not raw_request.app.state.enable_server_load_tracking:
return await func(*args, raw_request=raw_request, **kwargs)
return await func(*args, **kwargs)
raw_request.app.state.server_load_metrics += 1
try:
response = await func(*args, raw_request=raw_request, **kwargs)
response = await func(*args, **kwargs)
except Exception:
raw_request.app.state.server_load_metrics -= 1
raise