Web Development in Containers

I have been using asdf, a version manager for every tool (e.g. Node,js, Ruby, MySQL, you name it) for years. It was great since I could manage multiple versions of the same tool. I could switch between versions easily. I could even have different versions of the same tool in different projects.

There is nothing wrong with asdf. I thought to myself, “Is there a way to serve my web applications more efficiently?” Since I’ve been tinkering with Docker, maybe I could serve my web app with Docker and mount the working directory to the container so that any changes I make in the working directory would reflect in the container.

By serving the web app with a container, I do not have to go to the working directory, run and stop the server. As soon as I turn on my computer, the server is already running.

To do this, go to your working directory and create docker-compose.yaml and Dockerfile.

Dockerfile

1
FROM node:current-alpine3.19
2

3
WORKDIR /app
4
COPY . /app
5

6
RUN npm install
7
CMD ["npm", "run", "dev"]

1. FROM node:current-alpine3.19 - This line sets the node image with the tag current-alpine3.19 as the base image for this container.
2. WORKDIR /app - This line sets the working directory to /app.
3. COPY . /app - This line copies everything in the working directory to /app.
4. RUN npm install - This line installs the required dependencies in the container.
5. CMD ["npm", "run", "dev"] - This line serve the service.

By default, Next.js web applications are served on port 3000.

Docker Compose

1
services:
2
  webapp:
3
    build:
4
      context: .
5
      dockerfile: Dockerfile
6
    container_name: webapp
7
    volumes:
8
      - /path/to/your/webapp:/app
9
    restart: always
10
    ports:
11
      - 3000:3000

1. webapp: - This is the name of the service.
2. build: - This tells Docker to build the image.
3. context: . - This tells Docker to use the current directory you defined in Dockerfile above.
4. dockerfile: Dockerfile - This tells Docker to use the webapp.dockerfile we just defined above.
5. container_name: webapp - This is the name of the container, you could name whatever you like.
6. volumes: - This tells Docker to mount the working directory to the container. (Use pwd on your terminal if you are sure about the path you’re currently on).
7. restart: always - This tells Docker to restart the container if it crashes.
8. ports: - This tells maps the port 3000 in the container to the port 3000 on your computer.

To start the server, run docker compose up -d in your terminal. -d means detached mode, meaning the container will run in the background.

To make sure, go to your browser and access http://localhost:3000.

If there are some files that you do not wish to be copied into the container, do add .dockerignore into your codebase.

1
.astro
2
.next
3
...

If you need to import some environment variables, there are two ways you could consider.

1. Raw-dogging the environment values in the compose file like a real man.

1
services:
2
  webapp:
3
    build:
4
      context: .
5
      dockerfile: Dockerfile
6
    container_name: webapp
7
    environment:
8
      - DEBUG=TRUE
9
      - ...
10
    volumes:
11
      - /path/to/your/webapp:/app
12
    restart: always
13
    ports:
14
      - 3000:3000

2. Import .env file inside the compose file.

1
services:
2
  webapp:
3
    build:
4
      context: .
5
      dockerfile: Dockerfile
6
    container_name: webapp
7
    env_file:
8
      - .env
9
    volumes:
10
      - /path/to/your/webapp:/app
11
    restart: always
12
    ports:
13
      - 3000:3000

That way would import the .env file in the current working directory into the container.