TreeviewCopyright © Pengfei Ni all right reserved, powered by aleen42

GPU

Kubernetes支持容器请求GPU资源（目前仅支持NVIDIA GPU），在深度学习等场景中有大量应用。

在Kubernetes中使用GPU需要预先配置

  • 在所有的Node上安装Nvidia驱动，包括NVIDIA Cuda Toolkit和cuDNN等
  • 在apiserver和kubelet上开启--feature-gates="Accelerators=true"
  • Kubelet配置使用docker容器引擎（默认就是docker），其他容器引擎暂不支持该特性

使用方法

使用资源名alpha.kubernetes.io/nvidia-gpu指定请求GPU的个数，如

apiVersion: v1
kind: Pod
metadata:
  name: tensorflow
spec:
  restartPolicy: Never
  containers:
  - image: gcr.io/tensorflow/tensorflow:latest-gpu
    name: gpu-container-1
    command: ["python"]
    env:
    - name: LD_LIBRARY_PATH
      value: /usr/lib/nvidia
    args:
    - -u
    - -c
    - from tensorflow.python.client import device_lib; print device_lib.list_local_devices()
    resources: 
      limits: 
        alpha.kubernetes.io/nvidia-gpu: 1 # requests one GPU
    volumeMounts:
    - mountPath: /usr/local/nvidia/bin
      name: bin
    - mountPath: /usr/lib/nvidia
      name: lib
    - mountPath: /usr/lib/x86_64-linux-gnu/libcuda.so
      name: libcuda-so
    - mountPath: /usr/lib/x86_64-linux-gnu/libcuda.so.1
      name: libcuda-so-1
    - mountPath: /usr/lib/x86_64-linux-gnu/libcuda.so.375.66
      name: libcuda-so-375-66
  volumes:
    - name: bin
      hostPath:
        path: /usr/lib/nvidia-375/bin
    - name: lib
      hostPath:
        path: /usr/lib/nvidia-375
    - name: libcuda-so
      hostPath:
        path: /usr/lib/x86_64-linux-gnu/libcuda.so
    - name: libcuda-so-1
      hostPath:
        path: /usr/lib/x86_64-linux-gnu/libcuda.so.1
    - name: libcuda-so-375-66
      hostPath:
        path: /usr/lib/x86_64-linux-gnu/libcuda.so.375.66

$ kubectl create -f pod.yaml
pod "tensorflow" created

$ kubectl logs tensorflow
...
[name: "/cpu:0"
device_type: "CPU"
memory_limit: 268435456
locality {
}
incarnation: 9675741273569321173
, name: "/gpu:0"
device_type: "GPU"
memory_limit: 11332668621
locality {
  bus_id: 1
}
incarnation: 7807115828340118187
physical_device_desc: "device: 0, name: Tesla K80, pci bus id: 0000:00:04.0"
]

注意

  • GPU资源必须在resources.limits中请求，resources.requests中无效
  • 容器可以请求1个或多个GPU，不能只请求一部分
  • 多个容器之间不能共享GPU
  • 默认假设所有Node安装了相同型号的GPU

多种型号的GPU

如果集群Node中安装了多种型号的GPU，则可以使用Node Affinity来调度Pod到指定GPU型号的Node上。

首先，在集群初始化时，需要给Node打上GPU型号的标签

NVIDIA_GPU_NAME=$(nvidia-smi --query-gpu=gpu_name --format=csv,noheader --id=0)
source /etc/default/kubelet
KUBELET_OPTS="$KUBELET_OPTS --node-labels='alpha.kubernetes.io/nvidia-gpu-name=$NVIDIA_GPU_NAME'"
echo "KUBELET_OPTS=$KUBELET_OPTS" > /etc/default/kubelet

然后，在创建Pod时设置Node Affinity

kind: pod
apiVersion: v1
metadata:
  annotations:
    scheduler.alpha.kubernetes.io/affinity: >
      {
        "nodeAffinity": {
          "requiredDuringSchedulingIgnoredDuringExecution": {
            "nodeSelectorTerms": [
              {
                "matchExpressions": [
                  {
                    "key": "alpha.kubernetes.io/nvidia-gpu-name",
                    "operator": "In",
                    "values": ["Tesla K80", "Tesla P100"]
                  }
                ]
              }
            ]
          }
        }
      }
spec:
  containers:
    - image: gcr.io/tensorflow/tensorflow:latest-gpu
      name: gpu-container-1
      command: ["python"]
      args: ["-u", "-c", "import tensorflow"]
      resources:
        limits:
          alpha.kubernetes.io/nvidia-gpu: 2

使用CUDA库

NVIDIA Cuda Toolkit和cuDNN等需要预先安装在所有Node上。为了访问/usr/lib/nvidia-375，需要将CUDA库以hostPath volume的形式传给容器：

apiVersion: batch/v1
kind: Job
metadata:
  name: nvidia-smi
  labels:
    name: nvidia-smi
spec:
  template:
    metadata:
      labels:
        name: nvidia-smi
    spec:
      containers:
      - name: nvidia-smi
        image: nvidia/cuda
        command: [ "nvidia-smi" ]
        imagePullPolicy: IfNotPresent
        resources:
          limits:
            alpha.kubernetes.io/nvidia-gpu: 1
        volumeMounts:
        - mountPath: /usr/local/nvidia/bin
          name: bin
        - mountPath: /usr/lib/nvidia
          name: lib
      volumes:
        - name: bin
          hostPath:
            path: /usr/lib/nvidia-375/bin
        - name: lib
          hostPath:
            path: /usr/lib/nvidia-375
      restartPolicy: Never

$ kubectl create -f job.yaml
job "nvidia-smi" created

$ kubectl get job
NAME         DESIRED   SUCCESSFUL   AGE
nvidia-smi   1         1            14m

$ kubectl get pod -a
NAME               READY     STATUS      RESTARTS   AGE
nvidia-smi-kwd2m   0/1       Completed   0          14m

$ kubectl logs nvidia-smi-kwd2m
Fri Jun 16 19:49:53 2017
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 375.66                 Driver Version: 375.66                    |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  Tesla K80           Off  | 0000:00:04.0     Off |                    0 |
| N/A   74C    P0    80W / 149W |      0MiB / 11439MiB |    100%      Default |
+-------------------------------+----------------------+----------------------+

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

附录：CUDA安装方法

安装CUDA：

# Check for CUDA and try to install.
if ! dpkg-query -W cuda; then
  # The 16.04 installer works with 16.10.
  curl -O http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/cuda-repo-ubuntu1604_8.0.61-1_amd64.deb
  dpkg -i ./cuda-repo-ubuntu1604_8.0.61-1_amd64.deb
  apt-get update
  apt-get install cuda -y
fi

安装cuDNN：

首先到网站https://developer.nvidia.com/cudnn注册，并下载cuDNN v5.1，然后运行命令安装

tar zxvf cudnn-8.0-linux-x64-v5.1.tgz
ln -s /usr/local/cuda-8.0 /usr/local/cuda
sudo cp -P cuda/include/cudnn.h /usr/local/cuda/include
sudo cp -P cuda/lib64/libcudnn* /usr/local/cuda/lib64
sudo chmod a+r /usr/local/cuda/include/cudnn.h /usr/local/cuda/lib64/libcudnn*

安装完成后，可以运行nvidia-smi查看GPU设备的状态

$ nvidia-smi
Fri Jun 16 19:33:35 2017
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 375.66                 Driver Version: 375.66                    |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  Tesla K80           Off  | 0000:00:04.0     Off |                    0 |
| N/A   74C    P0    80W / 149W |      0MiB / 11439MiB |    100%      Default |
+-------------------------------+----------------------+----------------------+

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

results matching ""

    No results matching ""