Trust me it can get way more complex than that. RUN - run commands, as you would in a CLI.ĬMD - specifies the default command to run when starting a container from this image. dockerignore file depends on how the app is structured. You may need to omit files/directories which could be sensitive or auto-generated. WORKDIR - sets the working directory for any commands that come after it in the Dockerfile.ĬOPY - recursively copy files from the source path of the build context (scope of the Dockerfile on the host machine) to the destination path of the container filesystem. The choice of the image really depends on your requirements. I’m using an image from the official node repository. This image provides a runtime for the application. This is the basic structure of the Dockerfile:įROM - defines the base image to be used. When called to build the image, the docker daemon will look for that file in the defined scope, and then follow the instructions - hopefully the file exists, and has no errors. Moving on, the Dockerfile is a text file named Dockerfile (no file extension) placed within your application scope, generally at the root. In other words an image can exist without the container, while a container needs to run an image to exist - you get the idea. I think of an image as a “bundle” containing the environment required to run the application, and a container as the running instance of that image. Here’s my understanding of the two concepts: they are phases - a container is derived from an image. Container image! Container and Image are kind of the same, but fundamentally different. This is the set of instructions used to create the container image. Yup, it’s all about the Dockerfile (for the most part).

Nodejs dockerfile