xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-20 15:55:39 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Mis] remove duplicate engine status checks (#19647)

Browse Source 
Signed-off-by: googs1025 <googs1025@gmail.com>
This commit is contained in:
CYJiang 2025-06-17 23:17:38 +08:00 committed by GitHub
parent 4c8f64faa7
commit 5a1c2e15d8
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 3 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
vllm/v1/engine/core_client.py
View File

@ -794,7 +794,6 @@ class AsyncMPClient(MPClient):
request_type: EngineCoreRequestType,
request: Any,
engine: Optional[CoreEngine] = None) -> Awaitable[Any]:
self.ensure_alive()
if engine is None:
engine = self.core_engine
@ -1059,7 +1058,7 @@ class DPAsyncMPClient(AsyncMPClient):
self.reqs_in_flight.pop(req_id, None)
async def abort_requests_async(self, request_ids: list[str]) -> None:
if not request_ids:
if not request_ids or self.resources.engine_dead:
return
if len(request_ids) == 1:
@ -1077,9 +1076,8 @@ class DPAsyncMPClient(AsyncMPClient):
async def _abort_requests(self, request_ids: list[str],
engine: CoreEngine) -> None:
if not self.resources.engine_dead:
await self._send_input(EngineCoreRequestType.ABORT, request_ids,
engine)
await self._send_input(EngineCoreRequestType.ABORT, request_ids,
engine)
class RayDPClient(DPAsyncMPClient):