Installation Procedure for CI Pipeline(3.1)

Owned by Jenni Wallin
May 11, 2023
2 min read

The installation procedure for the CI pipeline involves the following steps:

  1. Unpack the tar archive (from Release Information) into a Git repository.

  2. Add a Helm chart for  under the Helm directory.

  3. The pipeline script currently supports deployment and test of the CI pipeline on a Kubernetes cluster locally with Minikube, Docker desktop, or to a local real Kubernetes cluster.
    If you plan to use a local real Kubernetes cluster, set up the following:
    1. Jenkins worker node should have an environment variable named ENV set to on-premise. That is, export ENV=on-premise.

    2. The pipeline.sh sets context to se2-kubernetes as default. If you want another context, change this setting in the script.

      The default is to install the CRD needed for . If the used cluster contains multiple installation only one can be control over the CRD.

      It is possible to add a boolean parameter to the pipeline named INSTALL_CRD. Then it can be controlled if the pipeline should install the CRD or not.

  4. Create the CI pipeline in Jenkins by doing the following:
    1. Create a new Item of type Pipeline.
    2. Under the section Pipeline in the configuration of the new item select Pipeline script from SCM.  
    3. Select your SCM and set Repository URL, credentials, and Branch.
    4. Point the Script Path to the jenkinsfile_customer in the Scripts folder, that is Scripts/jenkinsfile_customer.
  5. If images is located in a AWS ECR repository the following needs to be added to the Jenkins job.
    1. In the Configuration meny in Jenkins, click on This project is parameterised and add a String Parameter.

    2. The parameter needs to be named AWS_ECR_PROFILE.

      When the pipeline is running it will look up the credentials for the AWS Profile and addd that as a secret in the same namespace  will be deployed.

  6. If DTK MZPs are used the CI pipeline needs to push the DTK docker image to a Docker repository that can be used by the Kubernetes cluster. If DTK MZPs are not used this step can be ignored.
    1. Add a String parameter to the pipeline named DTK_REPO


The installation procedure contains the following sub section: