[Bug] 使用两个GPU运行“Qwen2-VL-2B”，启动后还没有任何请求就有一颗GPU自动满载运行

[jianliao]

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.

Describe the bug

使用两个GPU运行Qwen/Qwen2-VL-2B-Instruct，启动后还没有任何请求就有一颗GPU自动满载运行

Reproduction

lmdeploy serve api_server --backend pytorch --tp 2 Qwen/Qwen2-VL-2B-Instruct

Environment

sys.platform: linux
Python: 3.12.4 | packaged by Anaconda, Inc. | (main, Jun 18 2024, 15:12:24) [GCC 11.2.0]
CUDA available: True
MUSA available: False
numpy_random_seed: 2147483648
GPU 0,1: NVIDIA GeForce RTX 3090
CUDA_HOME: /home/jliao/miniconda3/envs/lmdeploy
NVCC: Cuda compilation tools, release 12.6, V12.6.20
GCC: gcc (Ubuntu 13.2.0-23ubuntu4) 13.2.0
PyTorch: 2.3.1+cu121
PyTorch compiling details: PyTorch built with:
  - GCC 9.3
  - C++ Version: 201703
  - Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v3.3.6 (Git Hash 86e6af5974177e513fd3fee58425e1063e7f1361)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - LAPACK is enabled (usually provided by MKL)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - CUDA Runtime 12.1
  - NVCC architecture flags: -gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_90,code=sm_90
  - CuDNN 8.9.2
  - Magma 2.6.1
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=12.1, CUDNN_VERSION=8.9.2, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=2.3.1, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=1, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF, 

TorchVision: 0.18.1+cu121
LMDeploy: 0.6.1+
transformers: 4.45.1
gradio: Not Found
fastapi: 0.115.0
pydantic: 2.9.2
triton: 2.3.1
NVIDIA Topology: 
	GPU0	GPU1	CPU Affinity	NUMA Affinity	GPU NUMA ID
GPU0	 X 	PHB	0-31	0		N/A
GPU1	PHB	 X 	0-31	0		N/A

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

Error traceback

N/A

[wangaocheng]

我遇到一个更奇怪的问题，我通过官方提供的Docker openmmlab/lmdeploy:latest安装了lmdeploy，我运行 lmdeploy serve api_server Qwen/Qwen2-VL-7B-Instruct --server-port 6001 --tp 2 的时候，docker 直接弹窗了。

[jianliao]

@wangaocheng 我看到这个issue了，建议你要么换成2B先试一试，要么或者不要用docker，直接跑在原生系统里试一试能否看到更多的错误信息。

[wangaocheng]

Windows 无法安装Triton，会出现 test failed!报错无法启动。

[grimoire]

正常的，nccl 等同步的 kernel

[jianliao]

@grimoire 请问这是目前的pytorch backend的局限性吗？有没有什么办法解决这个问题吗？

[grimoire]

空跑的 nccl kernel 应该没啥开销的，也不影响其他 stream 上 kernel 的使用的，也就显示的利用率高点，应该不是啥大问题吧

[jianliao]

空跑的 nccl kernel 应该没啥开销的，也不影响其他 stream 上 kernel 的使用的，也就显示的利用率高点，应该不是啥大问题吧

不是利用率高点，是一直保持在100%。。。

[grimoire]

理论上等待时用的 cuda core 很少的，就算100%也能同时在其他 stream 里跑别的 kernel 的。#2607 (review) 这里加了个 barrier，应该就不会 100% 了。

[github-actions]

This issue is marked as stale because it has been marked as invalid or awaiting response for 7 days without any further response. It will be closed in 5 days if the stale label is not removed or if there is no further response.

[github-actions]

This issue is closed because it has been stale for 5 days. Please open a new issue if you have similar issues or you have any new updates now.