Kubernetes에서 GPU 사용하기

sudo apt install -y apt-transport-https ca-certificates curl software-properties-common
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
sudo add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu focal stable"
apt-cache policy docker-ce
sudo apt install -y docker-ce

sudo docker run -it --runtime=nvidia --gpus all python:3.9 bash

docker: Error response from daemon: unknown or invalid runtime name: nvidia.
See 'docker run --help'

curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \
  && curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \
    sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
    sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list \
  && \
    sudo apt-get update

sudo apt-get install -y nvidia-container-toolkit

sudo nvidia-ctk runtime configure --runtime=docker

$ sudo cat /etc/docker/daemon.json
{
  "runtimes": {
      "nvidia": {
          "args": [],
          "path": "nvidia-container-runtime"
      }
  }

sudo systemctl restart docker

sudo docker run -it --runtime=nvidia --gpus all python:3.9 bash

$ nvidia-smi
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 470.199.02   Driver Version: 470.199.02   CUDA Version: 11.4     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  Tesla T4            On   | 00000000:00:05.0 Off |                  Off |
| N/A   27C    P8     8W /  70W |      0MiB / 16127MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+

sudo docker run -it --runtime=nvidia --gpus all python:3.9 bash

pip install torch==1.12.1+cu113 torchvision==0.13.1+cu113 torchaudio==0.12.1+cu113 -f https://download.pytorch.org/whl/torch_stable.html

>>> import torch
>>> torch.version.cuda
'11.3'

>>> torch.cuda.is_available()
True

import torch
>>> torch.cuda.is_available()
/usr/local/lib/python3.9/site-packages/torch/cuda/__init__.py:138: UserWarning: CUDA initialization: The NVIDIA driver on your system is too old (found version 11040). Please update your GPU driver by downloading and installing a new version from the URL: http://www.nvidia.com/Download/index.aspx Alternatively, go to: https://pytorch.org to install a PyTorch version that has been compiled with your version of the CUDA driver. (Triggered internally at ../c10/cuda/CUDAFunctions.cpp:108.)
  return torch._C._cuda_getDeviceCount() > 0
False

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: nvidia-device-plugin-daemonset
  namespace: kube-system
spec:
  selector:
    matchLabels:
      name: nvidia-device-plugin-ds
  updateStrategy:
    type: RollingUpdate
  template:
    metadata:
      labels:
        name: nvidia-device-plugin-ds
    spec:
      tolerations:
        - key: nvidia.com/gpu
          operator: Exists
          effect: NoSchedule
      # Mark this pod as a critical add-on; when enabled, the critical add-on
      # scheduler reserves resources for critical add-on pods so that they can
      # be rescheduled after a failure.
      # See https://kubernetes.io/docs/tasks/administer-cluster/guaranteed-scheduling-critical-addon-pods/
      priorityClassName: 'system-node-critical'
      containers:
        - image: nvcr.io/nvidia/k8s-device-plugin:v0.14.1
          name: nvidia-device-plugin-ctr
          env:
            - name: FAIL_ON_INIT_ERROR
              value: 'false'
          securityContext:
            allowPrivilegeEscalation: false
            capabilities:
              drop: ['ALL']
          volumeMounts:
            - name: device-plugin
              mountPath: /var/lib/kubelet/device-plugins
      volumes:
        - name: device-plugin
          hostPath:
            path: /var/lib/kubelet/device-plugins

sudo nvidia-ctk runtime configure --runtime=containerd
sudo systemctl restart containerd

$ cat <<EOF | kubectl apply -f -
apiVersion: v1
kind: Pod
metadata:
  name: gpu-pod
spec:
  restartPolicy: Never
  containers:
    - name: cuda-container
      image: nvcr.io/nvidia/k8s/cuda-sample:vectoradd-cuda10.2
      resources:
        limits:
          nvidia.com/gpu: 1 # requesting 1 GPU
  tolerations:
  - key: nvidia.com/gpu
    operator: Exists
    effect: NoSchedule
EOF

Test PASSED
Done

resources:
  limits:
    nvidia.com/gpu: 1 # requesting 1 GPU