How To Dockerize a Project

Knowing what a container is doesn't help until you can package your own app into one. Without a repeatable build-and-ship process, turning source code into a portable image and getting it into a registry is ad hoc and error-prone.

Solution: Write a Dockerfile (base image, copy code, install deps, set the start command), then build, tag, and push it to a registry.

To ship cargo through a port you never hand over a loose pile of goods; you seal them into one standardized shipping container so any crane and any ship can handle it the same way. What goes inside, and how it is packed, is written on the packing manifest taped to the door, and the sealed box is then stored in a bonded warehouse until a ship needs it. Dockerizing an app is identical: the Dockerfile is the packing manifest that says how to pack your source code into one container image, which you push to an image registry.

Take a real app (here a Node.js + Express API with server.js), write a Dockerfile, build an image, run it, and push it to Docker Hub.

Dockerizing an app is a four-step pipeline: write a Dockerfile, build an image, run a container, then push to a registry.

Graph legend — each node maps to a step that dockerizes the Express app:

These are the core instructions you combine to describe how an image is built and run.

  FROM        base image to start from
  WORKDIR     set working directory inside the image
  COPY/ADD    copy files from host into image
  RUN         execute a command at BUILD time (creates a layer)
  ENV         set environment variables
  EXPOSE      document the port the app listens on
  CMD         default command at RUN time (can be overridden)
  ENTRYPOINT  fixed command at RUN time (args appended)

A typical Node.js Dockerfile installs dependencies first, then copies the source, to make the most of layer caching.

FROM node:18-alpine
WORKDIR /app

# copy manifests first to leverage layer caching
COPY package*.json ./
RUN npm install --production

# then copy the rest of the source (server.js, routes/, ...)
COPY . .

EXPOSE 3000
CMD ["node", "server.js"]

Copying manifests before the source means dependency installs are cached and only rerun when dependencies actually change.

Graph legend — each node maps to a layer in the Express Dockerfile above:

Once the Dockerfile exists, you build the image, run it locally to test, then tag and push it to Docker Hub.

docker build -t express-api:1.0 .
docker run -d -p 3000:3000 --name express-api express-api:1.0
curl localhost:3000        # Express responds on port 3000

# tag + push to Docker Hub
docker login
docker tag express-api:1.0 mshaker/express-api:1.0
docker push mshaker/express-api:1.0

A .dockerignore file excludes files from the build context, just like .gitignore does for git.

node_modules
.git
Dockerfile
*.log
.env

Keeps the build context small and avoids leaking secrets.

CMD sets a default that arguments fully replace, while ENTRYPOINT sets a fixed command that arguments are appended to.

  CMD ["node","index.js"]          docker run img otherarg
       \__ fully replaced by "otherarg"

  ENTRYPOINT ["node"]              docker run img index.js
       \__ "index.js" appended -> node index.js

Turning your own app into a published image: write the Dockerfile, build, test locally, tag, then push.

Graph legend — each node maps to a real artifact/command for the Express app:

• Order Dockerfile steps from least- to most-frequently changed for caching.
• Use .dockerignore to shrink context & protect secrets.
• EXPOSE is documentation; -p actually publishes the port.

• [ ] Wrote a Dockerfile for a real app
• [ ] Built, ran, and curled it locally
• [ ] Pushed image to Docker Hub
• [ ] Added a .dockerignore