As we mentioned, the source-to-image build needs a builder image and you have to provide it each time when you are configuring such a build. The builder images contain scripts that are responsible for assembling and running the application. The assembling scripts will be run in phase 3 of the build algorithm, and the run script will be used as the start command of the resulting Docker image. During the build, the layer that contains the runnable application will be added on top of the builder image, the run script will be set as the image starting command, and the resulting image will be committed.
We know the basics of source-to-image builds, so now we can explain what we did when deploying our application in the last chapters. Let's start with the following command that we have invoked before running any builds:
oc create -f https...