EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

Optimize debug log position of PD abort request (#11090)

Browse Source 
Signed-off-by: Shangming Cai <csmthu@gmail.com>
This commit is contained in:
Shangming Cai
2025-10-03 23:07:02 +08:00
committed by GitHub
parent 03def5e3b1
commit 2c7f4ca2f2
1 changed files with 8 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
python/sglang/srt/managers/scheduler.py
View File

@@ -2597,31 +2597,31 @@ class Scheduler(
# Delete requests not in the waiting queue when PD disaggregation is enabled # Delete requests not in the waiting queue when PD disaggregation is enabled
if self.disaggregation_mode == DisaggregationMode.PREFILL: if self.disaggregation_mode == DisaggregationMode.PREFILL:
# Abort requests that have not yet been bootstrapped # Abort requests that have not yet been bootstrapped
for i, req in enumerate(self.disagg_prefill_bootstrap_queue.queue): for req in self.disagg_prefill_bootstrap_queue.queue:
logger.debug(f"Abort bootstrap queue request. {req.rid=}")
if recv_req.abort_all or req.rid.startswith(recv_req.rid): if recv_req.abort_all or req.rid.startswith(recv_req.rid):
logger.debug(f"Abort bootstrap queue request. {req.rid=}")
if hasattr(req.disagg_kv_sender, "abort"): if hasattr(req.disagg_kv_sender, "abort"):
req.disagg_kv_sender.abort() req.disagg_kv_sender.abort()
# Abort in-flight requests # Abort in-flight requests
for i, req in enumerate(self.disagg_prefill_inflight_queue): for req in self.disagg_prefill_inflight_queue:
logger.debug(f"Abort inflight queue request. {req.rid=}")
if recv_req.abort_all or req.rid.startswith(recv_req.rid): if recv_req.abort_all or req.rid.startswith(recv_req.rid):
logger.debug(f"Abort inflight queue request. {req.rid=}")
if hasattr(req.disagg_kv_sender, "abort"): if hasattr(req.disagg_kv_sender, "abort"):
req.disagg_kv_sender.abort() req.disagg_kv_sender.abort()
elif self.disaggregation_mode == DisaggregationMode.DECODE: elif self.disaggregation_mode == DisaggregationMode.DECODE:
# Abort requests that have not yet finished preallocation # Abort requests that have not yet finished preallocation
for i, decode_req in enumerate(self.disagg_decode_prealloc_queue.queue): for decode_req in self.disagg_decode_prealloc_queue.queue:
logger.debug(f"Abort prealloc queue request. {decode_req.req.rid=}")
if recv_req.abort_all or decode_req.req.rid.startswith(recv_req.rid): if recv_req.abort_all or decode_req.req.rid.startswith(recv_req.rid):
logger.debug(f"Abort prealloc queue request. {decode_req.req.rid=}")
if hasattr(decode_req.kv_receiver, "abort"): if hasattr(decode_req.kv_receiver, "abort"):
decode_req.kv_receiver.abort() decode_req.kv_receiver.abort()
# Abort requests waiting for kvcache to release tree cache # Abort requests waiting for kvcache to release tree cache
for i, decode_req in enumerate(self.disagg_decode_transfer_queue.queue): for decode_req in self.disagg_decode_transfer_queue.queue:
logger.debug(f"Abort transfer queue request. {decode_req.req.rid=}")
if recv_req.abort_all or decode_req.req.rid.startswith(recv_req.rid): if recv_req.abort_all or decode_req.req.rid.startswith(recv_req.rid):
logger.debug(f"Abort transfer queue request. {decode_req.req.rid=}")
if hasattr(decode_req.kv_receiver, "abort"): if hasattr(decode_req.kv_receiver, "abort"):
decode_req.kv_receiver.abort() decode_req.kv_receiver.abort()