Docker Tutorials

1. 安装Docker Engine

https://docs.docker.com/install/linux/docker-ce/ubuntu/

1.1 Docker Engine安装 (Ubuntu)

SET UP THE REPOSITORY

1. Update the apt package index:

$ sudo apt-get update

2. Add Docker’s official GPG key:

$ curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -

3. Use the following command to set up the stable repository.

$ sudo add-apt-repository \
    "deb [arch=amd64] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable"

INSTALL DOCKER ENGINE - COMMUNITY

Install the latest version of Docker Engine - Community and containerd

$ sudo apt-get update && sudo apt-get install docker-ce docker-ce-cli containerd.io

1.2 nvidia-docker安装（运行nvidia gpu需要安装）

https://github.com/NVIDIA/nvidia-docker

下面只适合docker version>=19.03

Ubuntu 16.04/18.04, Debian Jessie/Stretch/Buster

# Add the package repositories
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list

sudo apt-get update && sudo apt-get install -y nvidia-container-toolkit
sudo systemctl restart docker

CentOS 7 (docker-ce), RHEL 7.4/7.5 (docker-ce), Amazon Linux 1/2

distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.repo | sudo tee /etc/yum.repos.d/nvidia-docker.repo

sudo yum install -y nvidia-container-toolkit
sudo systemctl restart docker

2. 运行

2.1 启动docker daemon

Ubuntu:

$ sudo service docker start

CentOS:

$ sudo systemctl start docker

2.2 docker image 管理

查看所有images

docker images

删除docker image

docker rmi  <image>   ## image name or id

从Dockerhub或其他registry拉取一个image

docker pull  <image>  
docker pull  ubuntu:16.04

2.3 启动docker container

sudo docker run hello-world   #以root启动的docker daemon ， 启动docker image也需root 

启动一个ubuntu docker ，并进入bash

sudo docker run -it ubuntu bash

直接启动docker image

docker run -it <your-image>

nvidia runtime

docker run --runtime=nvidia -it <image>   # docker < 19.03
docker run --gpus all  -it <image>        # docker >=19.03

挂载 host dir

docker run -v  /host/dir:/container/dir <image>

端口映射

docker run -p hostport:container_port <image>

启动时设置环境变量

docker run -e RACV_DATA_PART=3  <image>

设置work_dir

docker run -w  /work_dir

使用主机的网络

docker run --network host xxx

2.3.1 container 查看/删除

查看所有container

docker ps -a
docker ps -a -q  ## 只看containerID

删除docker container

docker rm  container_name
docker rm $(docker ps -a -q)

2.4 查看container log

docker container logs  containerID

2.5 进入已经启动的docker container

docker exec -it containID  /bin/bash

2.6 stop container

docker container  stop containID
docker container  rm containID

2.7 debug container

https://medium.com/@betz.mark/ten-tips-for-debugging-docker-containers-cde4da841a1d

docker logs 

exec 只能在container running时可用，若container启动就崩溃，无法使用

3. Build docker image

https://docs.docker.com/engine/getstarted/step_four/

3.1 编写Dockerfile

https://docs.docker.com/engine/reference/builder/

FROM

DockerHub上一些官方image:

FROM library/ubuntu:14.04.4                                     
FROM nvidia/cuda:9.0-cudnn7-devel-ubuntu16.04   ##

公司自己的hub如：

FROM docker-registry.xxx.virtual/library/centos7:1.4
FROM docker-registry.xxx.virtual/docker/ubuntu:14.04.3

CMD

多条命令

CMD /etc/init.d/nullmailer start ; /usr/sbin/php5-fpm

ENV

设置进入bash的环境变量

RUN

https://docs.docker.com/engine/reference/run/

RUN pip install numpy

3.2 Build docker

$ docker build -t nginx_image PATH

PATH is required, context path, for ADD or COPY command
-t target name
-f dockerfile name
如:

$ docker build -t robotarm:0.2 . -f robotarm_algorithms/Dockerfiles/Dockerfile

4. push docker image 到registry

4.1 login

docker login  reg.xxx.com
sudo docker login reg.xxx.com   ## 有的登陆失败是由于没有用sudo

4.2 logout

docker logout  reg.xxx.com

4.3 镜像加tag

sudo docker tag maskrcnn-benchmark:0.2 reg.xxx.com/cv/maskrcnn-benchmark:0.2

同样的image 会多出一条记录：

REPOSITORY                          TAG     IMAGE ID        CREATED           SIZE    
maskrcnn-benchmark                  0.2     edfb8cb7c2a1    12 minutes ago    5.85GB    
reg.xxx.com/ocr/maskrcnn-benchmark  0.2     edfb8cb7c2a1    12 minutes ago    5.85GB   

4.4 推送镜像到registry的cv项目

sudo docker push  reg.xxx.com/cv/maskrcnn-benchmark:0.2

5. 其他

5.1 配置为无需sudo, run docker with non-root user(sudo)

https://docs.docker.com/install/linux/linux-postinstall/

1. Create the docker group.

$sudo groupadd docker

2. Add your user to the docker group.

$sudo usermod -aG docker $USER

3. Logout and log back in so that your group membership is re-evaluated.

5.2 Docker Hub -- share docker image

https://hub.docker.com/