查看linux 是否有gpu 机器

lspci | grep -i nvidia

查看 NVIDIA 驱动 是否安装

nvidia-smi

部署 gpu服务需要做以下主备

#参考 https://www.cnblogs.com/wuchangsoft/p/9767074.html

1.安装之前 ，必须先安装docker，docker-compose

2.配置nvdia 源

curl -s -L https://nvidia.github.io/nvidia-docker/centos7/x86_64/nvidia-docker.repo | sudo tee /etc/yum.repos.d/nvidia-docker.repo

3.查看可用的版本

yum search --showduplicates nvidia-docker

4.下载 nvidia-docker2

yum install nvidia-docker2-2.0.3 -y
如果报错一直执行这个命令

5.配置docker文件

vim /etc/docker/daemon.json

{
    "default-runtime": "nvidia",
    "runtimes": {
        "nvidia": {
            "path": "nvidia-container-runtime",
            "runtimeArgs": []
        }
    }
}

配置之前

配置之后

加一行 "default-runtime": "nvidia",

6.重启docker

systemctl restart docker

7.查看是否安装

yum list installed | grep nvidia

8. 开始部署 gpu服务

部署示例

参考官网：https://docs.docker.com/compose/gpu-support/

vi docker-compose.yaml

version: "3.7"
services:

#执照
  business-zhizhao-gpu:
    image: 镜像文件名称
    container_name: business-zhizhao-ocr
    command: [ "sh", "-c","./bin/docker_run.sh" ]
    #network_mode: "host"
    ports:
      - 8000:8000
    restart: always
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: "all"
              capabilities: [gpu]

注意 下面deploy是一组

启动
docker-compose up -d

查看服务状态
docker ps
查看日志
docker logs -f 23ds