[Bug] DeepSeek-R1-BF16 can't output with /v1/chat/completions on 4 node*8*A100

[zhaotyer]

Checklist

• 1. I have searched related issues but cannot get the expected help.
• 2. The bug has not been fixed in the latest version.
• 3. Please note that if the bug-related issue you submitted lacks corresponding environment info and a minimal reproducible demo, it will be challenging for us to reproduce and resolve the issue, reducing the likelihood of receiving feedback.
• 4. If the issue you raised is not a bug but a question, please raise a discussion at https://github.com/sgl-project/sglang/discussions/new/choose Otherwise, it will be closed.
• 5. Please use English, otherwise it will be closed.

Describe the bug

The model is loaded successfully and the model service has been started on 4 node8A100,
http://0.0.0.0:50050/generate Able to output content
http://0.0.0.0:50050/v1/chat/completions block

error is:
[rank16]:[E214 06:46:21.341483897 ProcessGroupNCCL.cpp:1785] [PG ID 2 PG GUID 3 Rank 16] Exception (either an error or timeout) detected by watchdog at work: 9301, last enqueued NCCL work: 9301, last completed NCCL work: 9300.
[rank16]:[E214 06:46:21.341531487 ProcessGroupNCCL.cpp:1834] [PG ID 2 PG GUID 3 Rank 16] Timeout at NCCL work: 9301, last enqueued NCCL work: 9301, last completed NCCL work: 9300.
[rank16]:[E214 06:46:21.341568027 ProcessGroupNCCL.cpp:630] [Rank 16] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
[rank16]:[E214 06:46:21.341649967 ProcessGroupNCCL.cpp:636] [Rank 16] To avoid data inconsistency, we are taking the entire process down.
[rank16]:[E214 06:46:21.343994164 ProcessGroupNCCL.cpp:1595] [PG ID 2 PG GUID 3 Rank 16] Process group watchdog thread terminated with exception: [Rank 16] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=9301, OpType=ALLREDUCE, NumelIn=129024, NumelOut=129024, Timeout(ms)=600000) ran for 600036 milliseconds before timing out.
Exception raised from checkTimeout at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:618 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x96 (0x7f8f6d56c446 in /usr/local/lib/python3.10/dist-packages/torch/lib/libc10.so)
frame #1: c10d::ProcessGroupNCCL::WorkNCCL::checkTimeout(std::optional<std::chrono::duration<long, std::ratio<1l, 1000l> > >) + 0x282 (0x7f8f2342a772 in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_cuda.so)
frame #2: c10d::ProcessGroupNCCL::watchdogHandler() + 0x233 (0x7f8f23431bb3 in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_cuda.so)
frame #3: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f8f2343361d in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_cuda.so)
frame #4: + 0x145c0 (0x7f8f6fd1d5c0 in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch.so)
frame #5: + 0x94ac3 (0x7f9012012ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #6: clone + 0x44 (0x7f90120a3a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

terminate called after throwing an instance of 'c10::DistBackendError'
what(): [PG ID 2 PG GUID 3 Rank 16] Process group watchdog thread terminated with exception: [Rank 16] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=9301, OpType=ALLREDUCE, NumelIn=129024, NumelOut=129024, Timeout(ms)=600000) ran for 600036 milliseconds before timing out.
Exception raised from checkTimeout at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:618 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x96 (0x7f8f6d56c446 in /usr/local/lib/python3.10/dist-packages/torch/lib/libc10.so)
frame #1: c10d::ProcessGroupNCCL::WorkNCCL::checkTimeout(std::optional<std::chrono::duration<long, std::ratio<1l, 1000l> > >) + 0x282 (0x7f8f2342a772 in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_cuda.so)
frame #2: c10d::ProcessGroupNCCL::watchdogHandler() + 0x233 (0x7f8f23431bb3 in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_cuda.so)
frame #3: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f8f2343361d in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_cuda.so)
frame #4: + 0x145c0 (0x7f8f6fd1d5c0 in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch.so)
frame #5: + 0x94ac3 (0x7f9012012ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #6: clone + 0x44 (0x7f90120a3a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

Exception raised from ncclCommWatchdog at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1601 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x96 (0x7f8f6d56c446 in /usr/local/lib/python3.10/dist-packages/torch/lib/libc10.so)
frame #1: + 0xe4271b (0x7f8f230a071b in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_cuda.so)
frame #2: + 0x145c0 (0x7f8f6fd1d5c0 in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch.so)
frame #3: + 0x94ac3 (0x7f9012012ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #4: clone + 0x44 (0x7f90120a3a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

Fatal Python error: Aborted

Thread 0x00007f88ddfff640 (most recent call first):

Reproduction

node1:
NCCL_P2P_LEVEL=NVL python3 -m sglang.launch_server --model-path /models/ --tp 32 --dist-init-addr 172.17.120.15:50030 --nnodes 4 --node-rank 0 --trust-remote-code --watchdog-timeout 21600 --mem-fraction-static 0.9 --host 0.0.0.0 --port 50050 --served-model-name atom --disable-cuda-graph
node2:
NCCL_DEBUG=TRACE NCCL_P2P_LEVEL=NVL python3 -m sglang.launch_server --model-path /models/ --tp 32 --dist-init-addr 172.17.120.15:50030 --nnodes 4 --node-rank 1 --trust-remote-code --watchdog-timeout 21600 --disable-cuda-graph --host 0.0.0.0 --port 50050
node3:

node4:
NCCL_DEBUG=TRACE NCCL_P2P_LEVEL=NVL python3 -m sglang.launch_server --model-path /models/ --tp 32 --dist-init-addr 172.17.120.15:50030 --nnodes 4 --node-rank 3 --trust-remote-code --watchdog-timeout 21600 --disable-cuda-graph --host 0.0.0.0 --port 50050

Environment

root@a100-15:/workspace# python3 -m sglang.check_env
WARNING 02-14 08:00:33 cuda.py:23] You are using a deprecated pynvml package. Please install nvidia-ml-py instead, and make sure to uninstall pynvml. When both of them are installed, pynvml will take precedence and cause errors. See https://pypi.org/project/pynvml for more information.
WARNING:mistral_common.tokens.tokenizers.multimodal:Warning: Your installation of OpenCV appears to be broken: module 'cv2.dnn' has no attribute 'DictValue'.Please follow the instructions at opencv/opencv-python#884 to correct your environment. The import of cv2 has been skipped.
/usr/local/lib/python3.10/dist-packages/pydantic/_internal/_config.py:345: UserWarning: Valid config keys have changed in V2:

• 'fields' has been removed
  warnings.warn(message, UserWarning)
  Python: 3.10.12 (main, Nov 20 2023, 15:14:05) [GCC 11.4.0]
  CUDA available: True
  GPU 0,1,2,3,4,5,6,7: NVIDIA A100-SXM4-80GB
  GPU 0,1,2,3,4,5,6,7 Compute Capability: 8.0
  CUDA_HOME: /usr/local/cuda
  NVCC: Cuda compilation tools, release 12.4, V12.4.131
  CUDA Driver Version: 550.90.07
  PyTorch: 2.5.1+cu124
  sglang: 0.4.2
  flashinfer: 0.1.6+cu124torch2.4
  triton: 3.1.0
  transformers: 4.48.2
  torchao: 0.8.0
  numpy: 1.26.4
  aiohttp: 3.9.3
  fastapi: 0.115.8
  hf_transfer: 0.1.9
  huggingface_hub: 0.28.1
  interegular: 0.3.3
  modelscope: 1.22.3
  orjson: 3.10.15
  packaging: 23.2
  psutil: 5.9.4
  pydantic: 2.10.6
  multipart: 0.0.20
  zmq: 25.1.2
  uvicorn: 0.34.0
  uvloop: 0.21.0
  vllm: 0.6.4.post1
  openai: 1.60.2
  anthropic: 0.45.2
  decord: 0.6.0
  NVIDIA Topology:
  GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 NIC6 NIC7 CPU Affinity NUMA AffinityGPU NUMA ID
  GPU0 X NV12 NV12 NV12 NV12 NV12 NV12 NV12 NODE NODE PXB PXB SYS SYS SYS SYS 0-23,48-71 0 N/A
  GPU1 NV12 X NV12 NV12 NV12 NV12 NV12 NV12 NODE NODE PXB PXB SYS SYS SYS SYS 0-23,48-71 0 N/A
  GPU2 NV12 NV12 X NV12 NV12 NV12 NV12 NV12 PXB PXB NODE NODE SYS SYS SYS SYS 0-23,48-71 0 N/A
  GPU3 NV12 NV12 NV12 X NV12 NV12 NV12 NV12 PXB PXB NODE NODE SYS SYS SYS SYS 0-23,48-71 0 N/A
  GPU4 NV12 NV12 NV12 NV12 X NV12 NV12 NV12 SYS SYS SYS SYS NODE NODE PXB PXB 24-47,72-95 1 N/A
  GPU5 NV12 NV12 NV12 NV12 NV12 X NV12 NV12 SYS SYS SYS SYS NODE NODE PXB PXB 24-47,72-95 1 N/A
  GPU6 NV12 NV12 NV12 NV12 NV12 NV12 X NV12 SYS SYS SYS SYS PXB PXB NODE NODE 24-47,72-95 1 N/A
  GPU7 NV12 NV12 NV12 NV12 NV12 NV12 NV12 X SYS SYS SYS SYS PXB PXB NODE NODE 24-47,72-95 1 N/A
  NIC0 NODE NODE PXB PXB SYS SYS SYS SYS X PIX NODE NODE SYS SYS SYS SYS
  NIC1 NODE NODE PXB PXB SYS SYS SYS SYS PIX X NODE NODE SYS SYS SYS SYS
  NIC2 PXB PXB NODE NODE SYS SYS SYS SYS NODE NODE X PIX SYS SYS SYS SYS
  NIC3 PXB PXB NODE NODE SYS SYS SYS SYS NODE NODE PIX X SYS SYS SYS SYS
  NIC4 SYS SYS SYS SYS NODE NODE PXB PXB SYS SYS SYS SYS X PIX NODE NODE
  NIC5 SYS SYS SYS SYS NODE NODE PXB PXB SYS SYS SYS SYS PIX X NODE NODE
  NIC6 SYS SYS SYS SYS PXB PXB NODE NODE SYS SYS SYS SYS NODE NODE X PIX
  NIC7 SYS SYS SYS SYS PXB PXB NODE NODE SYS SYS SYS SYS NODE NODE PIX X

Legend:

X = Self
SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)
NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node
PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU)
PXB = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge)
PIX = Connection traversing at most a single PCIe bridge
NV# = Connection traversing a bonded set of # NVLinks

NIC Legend:

NIC0: mlx5_0
NIC1: mlx5_1
NIC2: mlx5_2
NIC3: mlx5_3
NIC4: mlx5_4
NIC5: mlx5_5
NIC6: mlx5_6
NIC7: mlx5_7

ulimit soft: 1048576

[zhyncs]

Please use the latest version v0.4.3
https://docs.sglang.ai/start/install.html#method-1-with-pip