Security basics

A few habits that go a long way

You don't need to be a security expert to avoid the most common Docker mistakes. Here are the essentials, all building on things you already learned.

Pin your image versions

Back in the images chapter we mentioned tags. Relying on latest means your build can change unexpectedly when a new version is published. Pin a specific version so builds stay predictable:

# avoid
FROM node

# prefer
FROM node:22-alpine

Predictable builds are also more secure builds - you know exactly what you're running.

Never bake secrets into images

Passwords, API keys and tokens should never be written into a Dockerfile or copied into an image. Anyone who has the image can read them. Instead:

• Pass secrets as environment variables at run time (the -e flag or Compose environment).
• Keep them in a .env file that you exclude with .dockerignore and .gitignore.

Run as a non-root user

By default, processes inside a container run as root, the all-powerful user. If an attacker breaks into your app, running as root gives them more power. You can switch to a normal user with the USER instruction:

FROM node:22-alpine
WORKDIR /app
COPY . .
RUN addgroup app && adduser -S -G app app
USER app
CMD ["node", "server.js"]

• RUN addgroup ... adduser ... creates a non-root user (the exact command depends on the base image; this is the Alpine form).
• USER app switches to it, so everything after runs as that limited user.

Keep base images updated

Security fixes arrive in new versions of base images. Rebuilding periodically with an updated base pulls in those fixes. Combined with pinned versions, a good rhythm is: pin a version, and deliberately bump it now and then.

The leak that outlives the delete

We just said not to bake secrets in - here's the part people miss about why a quick fix won't save you. Deleting the file in a later step doesn't remove it. The secret still sits in the earlier layer, readable by anyone who pulls the image. Layers remember everything. So the rule isn't "delete it afterwards", it's "never let it into a layer in the first place".

FAQ

What are the most important Docker security practices for beginners?

Pin image versions instead of latest, keep secrets out of images (pass them at run time), and run containers as a non-root user with USER. Those three cover the most common mistakes.

Why shouldn't I put secrets in a Docker image?

Because anyone with the image can read them, and they persist in the image layers even if a later step deletes the file. Provide secrets via environment variables at run time, and exclude .env with .dockerignore.

How do I run a container as a non-root user?

Create a user in your Dockerfile and switch to it with the USER instruction, so the app no longer runs as all-powerful root. This limits the damage if the app is ever compromised.

Next steps after Docker basics

That's the course. You can now explain what Docker is, run and manage containers, build your own images with a Dockerfile, persist data with volumes, connect containers over networks, orchestrate multi-container apps with Compose, dockerize a real PHP app, and follow the practices that keep images small and secure.

From here, great next steps are learning about container orchestration in production (such as Kubernetes) and setting up automated image builds in your CI pipeline - but you now have the solid foundation they all build on.