[BugFix][EPD] Harden zmq_to_scheduler receiver failures; sync error info across TP (#31592)

Co-authored-by: siyu <liusy58@linux.alibaba.com>
This commit is contained in:
Zheng Wengang
2026-07-24 16:37:52 +08:00
committed by GitHub
co-authored by siyu
parent 58f417049d
commit 364b5f23e6
2 changed files with 105 additions and 45 deletions
@@ -226,11 +226,12 @@ class SchedulerRequestReceiver:
):
recv_reqs, abort_reqs = self.mm_receiver.process_waiting_requests(recv_reqs)
for req, error_msg, error_code in abort_reqs:
status_code = (
HTTPStatus.BAD_REQUEST
if error_code == 400
else HTTPStatus.INTERNAL_SERVER_ERROR
)
if error_code is None:
status_code = HTTPStatus.INTERNAL_SERVER_ERROR
elif isinstance(error_code, HTTPStatus):
status_code = error_code
else:
status_code = HTTPStatus(int(error_code))
prepare_abort(req, error_msg, status_code=status_code)
self.stream_output([req], req.return_logprob)
return recv_reqs