Building Images with a Dockerfile

What a Dockerfile is

A Dockerfile is a plain-text recipe that tells Docker how to build an image, one instruction per line. Each instruction creates a new read-only layer; layers are cached, so rebuilds only redo the steps that changed. Ordering instructions from least- to most-frequently-changing is the single biggest build-speed win.

A Dockerfile for a Node app

FROM node:20-alpine
WORKDIR /app

# Copy only manifests first so 'npm install' is cached
COPY package*.json ./
RUN npm install --production

# Now copy the rest of the source
COPY . .

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

Reading it line by line:

• FROM chooses a base image – here a small Alpine Linux image with Node 20.
• WORKDIR sets the working directory for later instructions.
• COPY package*.json then RUN npm install are done before copying the code, so dependencies are cached and only reinstalled when the manifest changes.
• EXPOSE documents the port; CMD is the default command run when a container starts.

Build and run it

docker build -t myapp:1.0 .
docker run -d -p 3000:3000 myapp:1.0

The -t flag tags the image with a name and version. The trailing . is the build context – the folder Docker sends to the daemon.

CMD vs RUN vs ENTRYPOINT

• RUN executes at build time and bakes the result into a layer (e.g. installing packages).
• CMD is the default command at run time and can be overridden on the command line.
• ENTRYPOINT sets a fixed executable; CMD then supplies its default arguments.

Key points

• A Dockerfile builds an image in cached layers – order instructions by change frequency.
• Copy dependency manifests and install before copying source to reuse the cache.
• docker build -t name:tag . builds and tags; RUN is build-time, CMD is run-time.