Up until now, we have set up a few elements:
- A GitHub repository with our code (
aggregator) - A continuous delivery pipeline in GCP for our Docker image that gets fired once we tag the code
- A Kubernetes cluster
- Jenkins connected to the preceding cluster
Now we are going to set up the continuous delivery pipeline for our code and the Kubernetes infrastructure. This pipeline is going to be actioned by a Jenkins job, which we will trigger manually.
You might be thinking that all that you have read about Continuous Delivery (CD) is about transparently shipping code to production without any human intervention, but here we are, with a few events that need manual steps in order to action the build. I have worked in some places where continuous delivery is triggered automatically by changes in the master branch of your repository, and after few incidents, I really believe that a manual trigger is a fair price to pay for having an enormous amount of control over...