Summary
The prefill worker (inkwell-copper-cn01) crashes with RuntimeError: Failed: CUDA error /build/DeepEP/csrc/deep_ep.cpp:112 'invalid resource handle' followed by an illegal memory access at line 278, causing the entire job to fail.
Error
RuntimeError: Failed: CUDA error /build/DeepEP/csrc/deep_ep.cpp:112 'invalid resource handle'
terminate called after throwing an instance of 'EPException'
what(): Failed: CUDA error /build/DeepEP/csrc/deep_ep.cpp:278 'an illegal memory access was encountered'
From: inkwell-copper-cn01_prefill_w0.out:1931 and inkwell-copper-cn01_prefill_w0.out:1946-1947
Config
The recipe passes these flags to prefill workers:
moe-a2a-backend: deepep
deepep-mode: normal
tp-size: 8
ep-size: 8
dp-size: 8
ep-dispatch-algorithm: dynamic
ep-num-redundant-experts: 32
mem-fraction-static: 0.75
disable-cuda-graph: true
Decode workers use different values (tp-size=32, ep-size=32, deepep-mode=low_latency, ep-dispatch-algorithm=static).
Job
- Job ID: 5051 (dsr1-fp8-1k1k-max-tpt-26203405870)
- Model: deepseek-ai/DeepSeek-R1 (DeepSeek-V3)
- Precision: FP8
- Topology: 2 prefill nodes (4 GPUs each) + 8 decode nodes (4 GPUs each)
- Container: lmsysorg_sglang_dev-cu13_20260521
Timeline
- 04:33:30 — Prefill workers complete first DeepGEMM warmup
- 04:34:21 — Prefill workers begin second DeepGEMM JIT session
- 04:35:30 — Prefill worker 0 crashes with SIGQUIT
- 04:35:38 — etcd connection failures begin (consequence)
- 04:35:44 — srtctl cleans up
Suggested Fix
- Add validation in the recipe to reject flag combinations that DeepEP cannot support — the "invalid resource handle" error suggests the CUDA resource configuration exceeds what the driver/hardware can provide
- Check DeepEP's max EP/TP limits — consider reducing ep-size from 8 to a validated maximum for prefill workers with deepep backend
- Consider reducing decode ep-size from 32 to 16 or 8 to stay within validated bounds
- Add a pre-flight check that validates DeepEP can initialize with the given topology before starting workers
Summary
The prefill worker (inkwell-copper-cn01) crashes with
RuntimeError: Failed: CUDA error /build/DeepEP/csrc/deep_ep.cpp:112 'invalid resource handle'followed by an illegal memory access at line 278, causing the entire job to fail.Error
From:
inkwell-copper-cn01_prefill_w0.out:1931andinkwell-copper-cn01_prefill_w0.out:1946-1947Config
The recipe passes these flags to prefill workers:
moe-a2a-backend: deepepdeepep-mode: normaltp-size: 8ep-size: 8dp-size: 8ep-dispatch-algorithm: dynamicep-num-redundant-experts: 32mem-fraction-static: 0.75disable-cuda-graph: trueDecode workers use different values (tp-size=32, ep-size=32, deepep-mode=low_latency, ep-dispatch-algorithm=static).
Job
Timeline
Suggested Fix