Containerize a Web Application using docker compose

#docker #services #containers #dockercompose

In this blog we will basically be containerizing a web application consisting of MySQL, phpmyadmin, frontend and backend as services using docker compose.
Also, we briefly get into how all those services are running and how multiple services make up an application running in multiple containers.

Note: This article does not provide concepts of docker. To know concepts of docker visit here.

Brief About Docker Compose

In short, docker compose helps to define and run docker applications in multiple containers. With docker compose we can define a yml file where we can define our services, networks, and the volumes required for our application to get it running. Basically, configuring, defining and managing all the necessary things required for our application becomes easy with docker compose.

And in this blog, we will also be defining and configuring simple services for our application through a compose file and run it using docker compose.

What is the application about (to be run with docker compose)?

The application we will be running with docker compose has 3 main services:

frontend service
backend service
mysql service (database service)

Also, we will be running the 4th service phpmyadmin which is used for managing and administering MySQL with a GUI.
This is a very simple application for demo, where in the mysql database there are lists of user information and through the frontend service we will request to the backend service to get all those users and then list them in our application UI. Like this, all those application services are dependent and collaborating with each other. We will be configuring these through docker compose.

Prerequisites

Before we start, lets make sure following things get installed into your system.

install docker
install docker compose

Note: I will be doing the whole setup with Ubuntu 22.04 LTS. Things might differ if you have a different operating system.

Getting Started

Since we have enough information about the application and the prerequisites are fulfilled. We can now jump to the section for defining and configuring our application with docker compose.

Initiate file structure

create a directory dockercompose
create a docker compose yml file dockercompose/compose.yml inside it.
Also create an empty testdump.sql file (I will mention the use of it later)

It should look something like this.

  📦dockercompose
      ┗ 📜compose.yml
      ┗ 📜testdump.sql

Defining and configuring services

Now inside the compose.yml let's define and configure the services for our application.

database service (mysql)

 version: "3.8"

 services:
    database:
      image: mysql
      environment:
        MYSQL_ROOT_PASSWORD: root
        MYSQL_DATABASE: demodb
      volumes:
        - ./testdump.sql:/docker-entrypoint-initdb.d/testdump.sql
      restart: always

Here, defining services in docker compose we use the keyword services. We have defined our service database configured with image:mysql and the environment variables required for it.
We can see it has a volumes keyword used here. We can mount local data to the docker container using the volumes keyword. In this context, we have mapped testdump.sql which is in our
local root directory which runs as an entrypoint when the container starts. It means when the database service is started in a container, the testdump.sql is the first thing which is run inside the container.
The entrypoint script will create 4 users in the database. Add some sql commands inside of testdump.sql:

    CREATE TABLE users(userid INT PRIMARY KEY AUTO_INCREMENT, fullname VARCHAR(60), email VARCHAR(60));
    INSERT INTO users (fullname, email) VALUES ('Sagar Gurung', 'sagar@example.com');
    INSERT INTO users (fullname, email) VALUES ('Bob Johnson', 'bob@example.com');
    INSERT INTO users (fullname, email) VALUES ('Eva White', 'eva@example.com');
    INSERT INTO users (fullname, email) VALUES ('Michael Brown', 'michael@example.com');
    ALTER USER 'root' IDENTIFIED WITH mysql_native_password BY 'root';
    flush privileges;

backend service
```
  backend:
    image: sagar4321/demo-server
    environment:
      PORT: 3001
      DBHOST: database
      DBPORT: 3306
      DBUSER: root
      DBNAME: demodb
      DBPASSWORD: root
    restart: always
    ports:
      - "3001:3001"
    depends_on:
      - database
```
Similarly, we define and configure our backend as above with the necessary image:sagar4321/demo-server (this image is built and pushed to docker hub which I created), environment and port 3001 (port where the server listens)
Here, we have another keyword depends_on which has the value database service. It sets the order in which services must start and stop. For the above case, the database service starts before the backend service starts.
frontend service
```
  frontend:
    image: sagar4321/demo-web
    environment:
      SERVER_URL: http://0.0.0.0:3001
    ports:
      - 8080:8080
    depends_on:
      - backend
```
Use a similar style of definition and configuration for the frontend service with image: sagar4321/demo-web (which I built and pushed to docker hub), environment and port 8080 (where the frontend service listen).
But here the frontend service depends on the backend service since it needs to be started before the frontend service.

phpmyadmin service

  phpmyadmin:
    image: phpmyadmin/phpmyadmin
    restart: always
    environment:
      PMA_HOST: database
      PMA_PORT: 3306
    ports:
      - "8081:80"
    depends_on:
      - database

Also, the same thing applies to defining and configuring the phpmyadmin service. It also depends on the database service, and listens to port:8081

Now the overall compose.yml file after combining all the services together looks something like this:

version: "3.8"

services:
  database:
    image: mysql
    environment:
      MYSQL_ROOT_PASSWORD: root
      MYSQL_DATABASE: demodb
    volumes:
      - ./testdump.sql:/docker-entrypoint-initdb.d/testdump.sql
    restart: always

  server:
    image: sagar4321/demo-server
    environment:
      PORT: 3001
      DBHOST: database
      DBPORT: 3306
      DBUSER: root
      DBNAME: demodb
      DBPASSWORD: root
    restart: always
    ports:
      - "3001:3001"
    depends_on:
      - database

  web:
    image: sagar4321/demo-web
    environment:
      SERVER_URL: http://0.0.0.0:3001
    ports:
      - 8080:8080
    depends_on:
      - server

  phpmyadmin:
    depends_on:
      - database
    image: phpmyadmin/phpmyadmin
    restart: always
    environment:
      PMA_HOST: database
      PMA_PORT: 3306
    ports:
      - "8081:80"

Yes, we have finally completed setting up our required services for our application through the compose file compose.yml
Now all it takes is the following command to run the whole group of services in multiple containers and get our application running.
From the directory dockercompose in the terminal just run the command:

docker compose -f compose.yml up

The command starts up all the services running together. Just wait a while so that all those services get up and running, then we can
see our application running in the browser. You can see different logs provided by different services in the console.

After a while we can access our frontend in our localhost which is running in port 8080. Accessing the frontend service shows the list of users
that we created in the database. The frontend service makes a request to the backend service. The backend service gets data from the database and
the data is returned to the frontend service which is rendered in the User Interface. This is how multiple services running in different containers can collaborate
with each other through configuration using docker compose. There are lots of thing we can do through docker compose but this is just the basics of it.

The application page you get should be something like this in the browser http://localhost:8080:

And that's all, I hope this blog was a useful for you on how we can run multiple services in multiple docker containers of an application using one docker compose file.