First, the environment
cat /etc/redhat-release
uname -r
Second, the installation
yum install docker
Uninstall docker
yum remove docker-*
Third, start
#启动
systemctl start docker
#查看状态
systemctl status docker
Error may start reading this https://www.jianshu.com/p/a62ffb13ada6
script installation
[root@vultr ~]# vim installDocker.sh
#!/bin/bash
# install docker shell
# made by if
# 0. 关闭防火墙
echo "0.使用脚本关闭防火墙..."
sudo systemctl stop firewalld
sudo systemctl disable firewalld
# 1. 使用脚本自动安装
echo "1.使用脚本自动安装..."
curl -fsSL get.docker.com -o get-docker.sh
sudo sh get-docker.sh --mirror Aliyun
# 2.启动 Docker CE
echo "2.启动 Docker CE..."
sudo systemctl enable docker
sudo systemctl start docker
# 3.测试 Docker 是否安装正确
echo "3.测试 Docker 是否安装正确..."
docker run hello-world
# 4.添加镜像加速器
echo "4.添加镜像加速器..."
echo "{
"registry-mirrors": [
"https://5ajk0rns.mirror.aliyuncs.com"
]
}" > /etc/docker/daemon.json
# 5.重新启动服务
echo "5.重新启动服务..."
sudo systemctl daemon-reload
sudo systemctl restart docker
sudo systemctl enable docker
[root@vultr ~]#sh installDocker.sh
Four, Docker basic commands
1. View the current mirror
[root@localhost ~]# docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
2. Search Mirror
execution docker search centos searches image from the dockerhub
docker search 镜像名字
3. Download image
The default is a foreign source, download will slow, mirroring suggested domestic warehouse
cat /etc/docker/daemon.json
{
"registry-mirrors": [ "https://registry.docker-cn.com"]
}
#重启docker
systemctl restart docker
Download image
docker pull centos #安装centos的镜像
docker pull nginx #安装nginx的镜像
4. Export mirror
docker save -o name Mirror Mirror
[root@localhost ~]# docker save -o nginx.tar nginx
[root@localhost ~]# ls
anaconda-ks.cfg nginx.tar
#需要将docker导出为tar，后面为镜像名称
5. Mirror introduced
docker load --input nginx.tar #使用input导入
docker load < nginx.tar #使用重定向导入
6. Remove Mirror
docker delete can be used later docker rmi plus docker's ID
Tip: If the image has been created a container, it will not be deleted
7. Delete container:
docker rm 容器名
或者使用
docker rm -f 容器名
删除所有容器
docker rm -f `docker ps -qa
The second vessel will be prompted to turn it off
8. Create a Startup container
docker run Mirror
docker container run -itd --name bs centos
#centos是镜像的名称，镜像的名称必须在选项的后面
#--name 容器的名称
#-t 让docker分配一个伪终端
#-i 让docker的标准输入打开{input}
#-d 后台分配一个终端
9. Check container
ps show is running -a container is not running the show
[root@localhost ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
1fca4596401c centos "/bin/bash" About a minute ago Up About a minute bs
[root@localhost ~]#
Start container
docker start bs #容器名启动
docker start bcededa4b82c #id号启动
Into the container
docker attach bs
Exit container
ctrl+d #关闭退出容器
ctrl+p+q #退出不关闭容器
Five, Docker Management Case
Tip: production scenarios is not in use docker attach into the vessel, we need to use nsenter this tool, this tool is included in util-linux package inside
[root@localhost ~]#yum install util-linux -y
#Centos7默认最小化已经安装
We nsenter can enter the container, but nsenter by pid into the container, so we need to know the pid container. We can get to by docker inspect pid
[root@localhost ~]# docker start abcdocker
abcdocker
[root@localhost ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
573227eb789b centos "/bin/bash" 25 minutes ago Up 1 second abcdocker
[root@localhost ~]# docker inspect -f "{{ .State.Pid }}" abcdocker
8603
[root@localhost ~]# nsenter -t 8603 -m -u -i -n -p
[root@573227eb789b /]#
docker inspect -f {{.State.Pid}} container vessel name or id
# Each container has .State.Pid, so this command in addition to the id of the container we need to find according to docker ps -a, all the other fixed the format
nsenter --target found above process --net --pid # enter the command id --mount --uts --ipc will enter into the container
Meaning of the parameters after the explanation nsenter instruction process id:
* –mount参数是进去到mount namespace中
* –uts参数是进入到uts namespace中
* –ipc参数是进入到System V IPC namaspace中
* –net参数是进入到network namespace中
* –pid参数是进入到pid namespace中
* –user参数是进入到user namespace中
We can get more parameters carried out by nsenter --help
We see the process into the container in
the boot process is nsenter
/ bin / bash is the process we run the container produced
-bash we use nsenter produced, so if we exit the container, the container will not quit, because -bash still running
Because every time you need to enter into the container that two commands, so we can write a script to get.
Script reads as follows:
[root@linux-node1 ~]# cat docker_in.sh
#!/bin/bash
# Use nsenter to access docker
docker_in(){
NAME_ID=$1
PID=$(docker inspect -f "{{ .State.Pid }}" $NAME_ID)
nsenter -t $PID -m -u -i -n -p
}
docker_in $1
Execution results are as follows
[root@linux-node1 ~]# chmod +x docker_in.sh
[root@linux-node1 ~]# ./docker_in.sh abcdocker
[root@f8c8c8156e26 /]# ps -ef
UID PID PPID C STIME TTY TIME CMD
root 1 0 0 Oct18 ? 00:00:00 /bin/bash
root 54 0 0 00:23 ? 00:00:00 -bash
root 67 54 0 00:23 ? 00:00:00 ps -ef
[root@f8c8c8156e26 /]#
We can not see into the container
[root@linux-node1 ~]# docker exec abcdocker ps -ef
UID PID PPID C STIME TTY TIME CMD
root 1 0 0 Oct18 ? 00:00:00 /bin/bash
root 85 0 0 00:28 ? 00:00:00 ps -ef
[root@linux-node1 ~]# docker exec abcdocker ls /
anaconda-post.log
bin
dev
Tip: You can use exec parameters, do not view the contents into the container
We can also use the exec into the docker container
[root@linux-node1 ~]# docker exec -it abcdocker /bin/bash
But it is better to use less exec, there might be some unexpected effects caused by container
original http://www.dhhblog.com/?p=539
