You will create a Toolchain Service that contains a Delivery Pipeline. The Delivery Pipeline will contain the Tekton resources required to execute automated build and deploy steps when a trigger is clicked.
Navigate to the tekton-toolchain GitHub repository and make a fork. I will refer to this fork as https://github.com/powellquiring/tekton-toolchain in this tutorial, when you see this replace it with your fork.
If warnings or errors Continuous Delivery service required... are displayed you can live with the warnings or create a Continuous Delivery service (Lite plan) for the region.
This first part is about the simplest pipeline that can be configured and added to a pipeline service of a toolchain.
Visit the IBM Cloud in your browser. In the hamburger menu in the upper left choose DevOps.
- Create a toolchain, Build your own toolchain, in the context of this toolchain:
- Add tool GitHub
- GitHubServer: GitHub (https://github.com)
- repository type: existing
- Repository URL: this repo or your fork: https://github.com/powellquiring/tekton-toolchain
- Add tool Delivery Pipeline, name: whatever, Pipeline type: Tekton
As you can see a toolchain is a landing page that holds integrated tools. You have integrated two tools: GitHub and a Delivery Pipeline.
Click the Delivery Pipeline to open. Notice the context is in the Delivery Pipeline service just created. The Configuration page is likely displayed, if not click Configure Pipeline (if you do not see a Configure Pipeline button maybe you did not select Tekton when you created it)
- Select the Definitions panel. Definitions are going to define the set of tekton files that will contribute to this Delivery Pipeline Service. Contribute all of the files in the lab1-simple path in your clone (I show my clone below):
| Repository | Branch | Path |
|---|---|---|
| https://github.com/powellquiring/tekton-toolchain | master | lab1-simple |
- Worker: The default works great and for me it was: (Beta)IBM Managed workers in DALLAS
- Triggers:
- Add trigger
- Manual Trigger
- EventListener: the-listener
Save and then Close and you are back to the delivery pipeline: click Run Pipeline > Manual Trigger. The View the PipelineRun link will take you to the results.
This demonstrates some cool stuff. All of the tekton files in the github lab1-simple. In this case it was just the tekton.yaml file but all files with a .yaml or .yml extension would have been read by the Delivery Pipeline service.
- The name of the trigger in the drop down is the same as the EventListener in the lab1-simple/tekton.yaml file:
apiVersion: tekton.dev/v1beta1
kind: EventListener
metadata:
name: the-listener
...
Click on the log entry created and you will see xx1 which matched the task name in the pipeline:
apiVersion: tekton.dev/v1beta1
kind: Pipeline
metadata:
name: pipeline
spec:
tasks:
- name: xx1
And you see the task steps that were executed. No surprises. The steps are commands executed in the associated container and give you a feel for the environment. The ubuntu container was supplied by hub.docker.com
apiVersion: tekton.dev/v1beta1
kind: Task
metadata:
name: the-task
spec:
steps:
- name: echo
image: ubuntu
command:
- echo
args:
- "01 version"
- name: lslslash
image: ubuntu
...
Note that the last few steps show that the file system is not preserved between steps.