Prior to recreating the Continuous Integration pipeline job, we need a S3 bucket for our YUM repository. Create a bucket (unless you've kept the old one around), update the demo-app/Jenkinsfile script accordingly then commit and push Git changes upstream.
Refer to the Setting up the pipeline steps from the previous chapter to create the Continuous Integration job. Let us call it demo-app this time around. The script path remains the same (https://git-codecommit.us-east-1.amazonaws.com/v1/repos/demo-app).
You should now have this:
The pipeline is going to fail as we do not have our YUM repository configured yet:
The repository contents have already been uploaded to S3 by this first job run. Now we need to update the salt/states/yum-s3/files/s3.repo file with the S3 URL and set the repository to enabled. Commit and push the Salt changes to the Git repository, then pull and apply on the Jenkins node.
A subsequent pipeline run takes us a step further...