When you build an image from a Dockerfile, Docker first checks if the base image exists locally or pulls it from Docker Hub if needed. It then sets up a working environment within the container, copies necessary files, installs any dependencies, and sets the command that launches the application.

Dockerfiles utilize several key commands:

• FROM initiates the build process by setting the base image.
• WORKDIR establishes the directory for subsequent instructions.
• COPY and ADD are used to transfer files into the container, with COPY being preferred for local files.
• RUN executes commands within the container.
• CMD and ENTRYPOINT dictate the default command or entry point for the container.

Each Dockerfile command adds a layer to the Docker image, with Docker caching these layers to expedite future builds. If a command hasn't changed since the last build, Docker reuses the cached layer. This caching mechanism underscores the importance of ordering Dockerfile instructions wisely to avoid unnecessary rebuilds.

Three best practices for Dockerfiles are:

1. Using a .dockerignore file to prevent unnecessary files from bloating the build context, similar to .gitignore in Git repositories.
2. Minimizing the number of image layers by consolidating commands, which can speed up startup times.
3. Strategic ordering of instructions to maximize cache efficiency, placing less changeable instructions earlier.

For example, copying a package.json file and running npm install before adding application code helps preserve the cache for dependency installation when only the application code changes.

Understanding these elements and best practices helps you build efficient Docker images and optimize build times, essential for effective Docker use.

https://kodekloud.com/blog/how-dockerfile-works/