Reprinted from https://juejin.cn/post/7050304120082661407

It's 2022. We'll spend some time learning docker. We won't be fooled

Haven't you learned about docker yet? It doesn't matter. Let's simulate a few scenarios and find out what it is and when we need it in two minutes

Then we take the remaining seven or eight minutes and use docker to deploy vue2 and vue3 projects at the same time. Yes, that's so fast!

1 what is docker

Scenario simulation 1: single machine

In the company, you are diligent and quickly recognized by the leader. On this day, he asked you to take over the following new projects. You need to install them on your computer

You can see that there are many different versions of nodejs and mysql: OK, it's small for me

Scenario simulation 2: multi machine

At this time, the company has three new front-end interns and newcomers. The leader wants you to install these projects on their local computers

Due to the company's history and other reasons, they can only be equipped with computers with different operating systems: win7, win10, OSX, etc

Now you need to install these items on different operating systems of several computers. See the figure

Witty you find that one of your heads is bigger than two

"Installation is troublesome and time-consuming. When new people do projects, they have to switch matching versions for different projects. It is easy to make mistakes, which also pollutes the host environment and the host environment. When they think that there are more projects to do in the future, someone may already want to run away."

After awesome installation, you can do that. The leader has asked you to deploy these projects on the two small servers of the company's internal network, and do the public development and testing environment. Oh, these two servers are linux. Are you ready to start Baidu installing node and mysql in linux? And you still have to install multiple versions. What if the linux systems of the two servers are different?

Now it's your turn to carry the bucket

3 docker exit

Don't panic. Here comes docker, which can help you smooth out the differences in application installation between different systems

Wait, why not use a virtual machine?

Because the virtual machine starts slowly, consumes a lot of resources and occupies a lot of resources, the migration / expansion on different systems is more complex

But docker doesn't. It starts in seconds, takes up less resources, smells good, and is easy to copy across platforms

After the configuration is written, it installs different versions of nodej, msyql, nginx, etc. on the computer at the same time with a command, and runs independently of each other. At this time, you don't need to switch back and forth manually

Isn't the above problem easily solved? Are you interested in docker now

2 docker Foundation

Now let's start to understand the basics of docker

docker is divided into several concepts: image, container and warehouse

Image: it is like the system disk or system image file we need when installing. Here, it is responsible for creating the docker container. There are many official ready-made images: node, mysql, mono and nginx, which can be downloaded from the remote warehouse

Container: it can be compared to a mini system, such as one that only installs mysql5 7 is the smallest linux system. Of course, you can also install MySQL and node in the same container if you like. Remember * *, containers and containers, containers and hosts are isolated from each other**

Warehouse: the warehouse is like a github. We can create an image and then push it to the cloud warehouse, or we can pull and download the image from the warehouse

3 installation

The installation of docker is very simple. Both win and osx are installed with graphical interfaces. linux also has a few lines of commands. Now the m1 chip series of mac also supports it. Let's skip the installation steps first. Be fast! Finish it first!

ps: installing mysql on the docker of m1 chip requires a little configuration

After installation, run the code below to view

docker -v
 Copy code

4 actual combat: deploy vue2 and vue3 projects

After the docker is installed, let's go to the actual combat and rub our hands

We want the computer to run multiple versions of nodejs10 and nodejs12 at the same time

ps: let's get started quickly. Let's put aside the installation of different versions of mysql for the time being

4.1 preparation of vue2 and vue3 projects

Do it together first, and we'll explain later

Now create a new file for our project: name it: my repository

Install vue2+webpack project

# The 0 command line goes to the location of the folder: 
cd /The specific file path of your computer/my-repository

# 1. Now install Vue cli
npm install -g @vue/cli

# 2. Check whether Vue cli is installed successfully
vue --version
#This is @ vue/cli 4.5.15

# 3. Use Vue cli to quickly create a project. The installation options are as follows:
# > ❯ Default ([Vue 2] babel, eslint) 
# >❯ npm package management
vue create my-app-vue2

Copy code

Install vue3+vite project

#Install the latest version of vite first
npm init vite@latest

# Create vue3 project
npm init vite@latest my-app-vue3 --template vue

Copy code

//vite needs to turn on network access
//vite.config.js open host
export default defineConfig({
  plugins: [vue()],
+  server: {
+    host: '0.0.0.0',
+  },
});
Copy code

#After the installation, our directory is like this
my-repository
├── my-app-vue2
│   ├── public
│   └── src
│       ├── assets
│       └── components
└── my-app-vue3
    ├── public
    └── src
        ├── assets
        └── components
 Copy code

4.2 create and run docker container

# 0 first enter the folder where we just installed the vue project
cd my-repository

# 1 execute pwd to obtain the absolute directory of the current folder on the computer
pwd
# /Users/eric/my-repository

# 2 run to create docker container 1: host vue2+webpack+nodejs10
docker run -it -d --name myvue2 --privileged -p 8081:8080 -v  /Users/eric/my-repository/my-app-vue2:/app/vue node:10.16.2 /bin/bash -c "cd /app/vue && node -v && npm install && npm run serve"

# 3 run to create docker container 2: host vue3+vite+nodejs12
docker run -it -d --name myvue3 --privileged -p 8080:3000 -v /Users/eric/my-repository/my-app-vue3:/app/vue node:12.22.6 /bin/bash -c "cd /app/vue && node -v && npm install && npm run dev"

#Check the container operation after successful operation
docker ps -a

Copy code

Appears after successful operation

We can see the startup status, port mapping and container name of the container

Open the browser and visit localhost:8080 and localhost:8081 to see

If there is an error, see the third point below: [debugging] (####3 debugging): run the following command to check the reason

docker logs -f container_id/containe_name
 Copy code

What is the code of docker run xxxxx above? Now let's smooth it out

First, the docker run can be used to create a container for simultaneous startup and operation

Let's start with a new line: when the shell script is too long, we can use "\" to divide a line of command into multiple lines

docker run \
-it \
-d \ 
--name myvue2 \
--privileged \
-p 8081:8080 \
-v /Users/eric/my-repository/my-app-vue2:/app/vue \
node:10.16.2 \
/bin/bash -c "cd /app/vue2 && node -v && npm install && npm run serve"
Copy code

Here, we can use the docker run command to download the image - > create a container through the image - > start and run the container

Parameter resolution:

Running diagram of docker run

After the above code runs successfully, our computer will have two docker containers running independently of each other

4.3 commissioning

Common debugging commands 1

# After running, press ctrl + c to exit
docker logs -f contianer_name/container_id
 Copy code

Of course, we can use this command to view the information output by the terminal when compilation is in progress or errors or even exits in the container

🌰 : After running successfully, view the real-time output information of npm run serve of myvue container on the terminal

#View the terminal output information of docker container 
docker logs -f myvue2
 Copy code

Common debugging commands 2

# Print out the container's port mapping, directory mount, network, and so on
docker inspect myvue2
 Copy code

5 common operation commands

The column of common operation command table needs to be collected first

Mirror operation command

# Search image
docker search [images_name:tag]

# Download Image (: specify version)
docker pull [images_name:tag]

# View locally downloaded images
docker images

# Build your own image
# Build an image according to the path or url of the dockerfile
 docker build [OPTIONS] PATH|URL|-

# View the construction history of the mirror
docker history [images_name]

# delete mirror
# You need to delete the container based on this image first
 docker rmi [images_name]
Copy code

Container operation command

# Viewing running containers
# You can view container ID, basic image, container name, running status, port mapping, etc
docker ps

# View all containers: including stopped
docker ps -a

# View container information
# For example, port number mapping and directory mounting
docker inspect [images_name/images_id]

# Start and stop containers
docker start/stop [container_name/container_id]

#  Restart container
#  Usage scenario instance:
#  When adding a new npm package dependency requires recompilation, use restart to run the compilation
#  After the configuration of nginx container is updated, it needs to be restarted to take effect
docker restart [container_name/container_id]

# Enter container
# ps: some containers do not have bash and need to be changed to / bin/sh, such as mysq and mongodb
# Exit the container and enter exit
docker exec -it [container_name/container_id] /bin/bash

# Delete container
# Only when the container is stopped can it be deleted
docker rm [container_name/container_id]

# Container host file copy
# Copy container files to host
docker cp [container_id/container_name] : [File directory] [Host Directory]

# Copy the directory of the host to the container
docker cp [Host Directory] [container_id/container_name] : [File directory]


Copy code

6 Advanced

If there is no suitable image, we usually use Dockerfile to build a custom image

Did you find that the above docker run can only create and start one docker container. We can use docker compose to start multiple containers at one time. It is often used to install multiple services under a single machine

Please update later. If you are interested, you can also see my deployment with docker first Jenkins Automated Deployment CI/CD environment There is also the use of docker compose