Understanding Docker Images and How to Build Them

Docker images are executable filesystem packages that contain all code, libraries, environment variables, and configuration files needed to run an application. They are built in layers: the bottom bootfs (never directly interacted with), the rootfs (a base operating system), and additional layers added by the user.

These layers are combined using a union mount so the file system appears as a single, unified view, while remaining read‑only to guarantee environment consistency. When a container runs, Docker adds a writable layer on top of the image, allowing file modifications without altering the underlying image.

To build a custom image, first prepare your project files, then create a Dockerfile that specifies a base image, sets metadata, defines a working directory, copies project files, sets environment variables, exposes ports, and defines the container’s entrypoint command. You can also add a .dockerignore file to exclude unnecessary files, similar to .gitignore.

Build the image with the command:

docker build --build-arg NODE_ENV=develop -t docker-demo .

The -t flag tags the image, and --build-arg passes build‑time variables to the Dockerfile. After the build completes, run a container from the image with:

docker run --name demo -it -p 9999:8888 docker-demo

The --name option gives the container an alias, and -p maps host port 9999 to container port 8888, allowing you to access the web service at localhost:9999. If image download is slow, a registry accelerator such as DaoCloud can be used.

In summary, Docker images are read‑only, layered filesystems, and containers provide a writable runtime layer; mastering Dockerfile creation and image building is essential for modern cloud‑native development.