About

This repo provides Sample Azure DevOps Automation for using Azure DevOps on OpenShift. It's styled around the steps discussed here

• Azure DevOps with Managed OpenShift
• Hosting an Azure Pipelines Build Agent in OpenShift It uses Terraform and Helm to deploy the example described above.

Repo contains example Azure Pipeline Deployment and automation

• Pipeline Example 1

Requirements

• Terraform >=v1.10.0
• Helm
• OpenShift OC cli
• Azure Portal
• Github
• Bash(Some of the Terraform steps call out to Bash Scripts)
• Was run and tested with OCP 4.15

General Steps

• This example requires the OpenShift Image Registry is publically exposed.

  oc patch configs.imageregistry.operator.openshift.io/cluster --patch '{"spec":{"defaultRoute":true}}' --type=merge

• Create an Azure DevOps Organization

• Create Azure DevOps Personal Access Token. Automation was tested with a token that had full access.

• Create your Github Personal Token.You can use fine-grained tokens restricted to the repos you will fork for the examples below.

Deploy Pipeline Example 1

Pipeline Example 1 will
1 Use Terraform to deploy an Azure DevOps Build Agent on OpenShift
2 Will create the required Azure Infrastucture on Azure DevOps(Pipeline, Registry and access to OpenShift)
3 When pipeline is started, Pipeline will build a Dotnet image, push it into the internal OpenShift image registry and deploy a sample application using that image.

Steps to deploy Example 1

• Fork Sample Github Repository.Please edit the azure-pipelines.yaml file in your forked repo, change the name of the devops pool to a name of your choice. Example - azure-pipelines.yaml- Example used "AzurePipeline" . The automation expects to create the pool and will fail if the pool already exists. Pool name cannot be "Default"

• Automation requires that you are logged into OpenShift Cluster before running and that you provide the active kubeconfig as part of the steps. Example method to be run inside forked repo folder

  mkdir ./.kube
  touch ./.kube/config
  export KUBECONFIG=$PWD/.kube/config
  oc login

• Export the variables needed for automation

  TF_VAR_AZP_URL = AZURE DevOps Org URL
  TF_VAR_AZP_TOKEN = Azure DevOps Personal Access Token
  TF_VAR_AZP_POOL = Azure DevOps Pool name set in azure-pipelines.yaml above
  TF_VAR_GITHUB_REPO_NAME = Your Github Repo
  TF_VAR_GITHUB_REPO_BRANCH = Github Branch for your Repo
  TF_VAR_GITHUB_AZURE_PIPELINE_PATH = Path in Github Repo for azure-pipelines file.
  TF_VAR_AZDO_GITHUB_SERVICE_CONNECTION_PAT = Github Personal Token
  TF_VAR_KUBE_CONFIG_PATH = Path to active kubeconfig file

  Example Export

  export TF_VAR_AZP_URL=https://dev.azure.com/YourOrg
  export TF_VAR_AZP_TOKEN=.......
  export TF_VAR_AZP_POOL="AzurePipeline"
  export TF_VAR_GITHUB_REPO_NAME="MoOyeg/azure-pipelines-openshift"
  export TF_VAR_GITHUB_REPO_BRANCH="master"
  export TF_VAR_GITHUB_AZURE_PIPELINE_PATH="azure-pipelines.yml"
  export AZDO_PERSONAL_ACCESS_TOKEN=${TF_VAR_AZP_TOKEN}
  export AZDO_ORG_SERVICE_URL=${TF_VAR_AZP_URL}
  export TF_VAR_AZDO_PERSONAL_ACCESS_TOKEN=${TF_VAR_AZP_TOKEN}
  export TF_VAR_AZDO_ORG_SERVICE_URL=${TF_VAR_AZP_URL}
  export TF_VAR_AZDO_GITHUB_SERVICE_CONNECTION_PAT=.........
  export KUBE_CONFIG_PATH=${KUBECONFIG}
  export TF_VAR_KUBE_CONFIG_PATH=${KUBECONFIG}  

• Run automation to deploy pipeline example

    cd ./pipeline-example1-terraform
    terraform init --upgrade
    terraform apply -auto-approve

• Confirmation

  • Go to the created project in Azure Portal
    

  • Select the create pipeline object

  • Run the Pipeline

  • Check Pipline Status after uccessful Run

  • P.S There is a small error that might show up. "Error: no names or ids specified. Havent yet figured out what causes it. Pipeline should still run successfully.

CleanUp Example 1

terraform destroy

Helm Only for Example 1

If you do not want the end-to-end terraform install and want do the examples manually. Run the same steps except the terraform from above.Helm can be used to deploy the rest.

helm install azure-build-agent-openshift \
./charts/azure-build-agent-openshift/ \
--set azp_url=${AZP_URL} \
--set azp_token=${AZP_TOKEN} \
--set azp_pool=${AZP_POOL} \
--set serviceAccount.name="azure-build-agent-openshift-sa" \
--create-namespace \
--namespace azure-build

helm install azure-pipeline-openshift \
./charts/azure-devops-pipeline/ \
--create-namespace \
--set serviceAccount.name="azure-sa" \
--set serviceAccount.secretname="azure-sa-devops-secret" \
--set buildNamespace="azure-build" \
--set deploy_arogcd_app="false" \
--namespace ado-openshift

Cleanup for Helm Only

helm uninstall azure-build-agent-openshift --namespace azure-build
helm uninstall azure-pipeline-openshift --namespace ado-openshift

Deploy Pipeline Example 2

Pipeline Example 2 will
1 Install OpenShift gitops
1 Use Terraform to deploy an Azure DevOps Build Agent on OpenShift and an ArgoCD Application
2 Will create the required Azure Infrastucture on Azure DevOps(Pipeline, Registry and access to OpenShift)
3 When pipeline is started, Pipeline will build a react image, push it into the internal OpenShift image registry. Pipeline will update the git repo with the details of the new image which the ArgoCD application should auto sync.

Steps to deploy Example 2

• Complete the General Requirement Steps

• Destroy Example 1 if installed before proceeding

• Fork Sample Github Repository.Please edit the your azure-pipelines.yaml file, change the name of the devops pool to a name of your choice.

• Automation requires that you are logged into OpenShift Cluster before running and that you provide the active kubeconfig as part of the steps. Example method to be run inside forked repo folder

  mkdir ./.kube
  touch ./.kube/config
  export KUBECONFIG=$PWD/.kube/config
  oc login

• Install OpenShift Gitops

  oc adm new-project openshift-gitops
  oc apply -k https://github.com/redhat-cop/gitops-catalog/openshift-gitops-operator/operator/overlays/latest

• Give OpenShift Gitops permission for our soon to be created application namespace oc adm policy add-cluster-role-to-user admin system:serviceaccount:openshift-gitops:openshift-gitops-argocd-application-controller -n ado-openshift

• Export the variables needed for automation TF_VAR_AZP_URL = AZURE DevOps Org URL TF_VAR_AZP_TOKEN = Azure DevOps Personal Access Token TF_VAR_AZP_POOL = Azure DevOps Pool name set in azure-pipelines.yaml above TF_VAR_GITHUB_REPO_NAME = Your Github Repo(For Example 2 should be your fork of this repo) TF_VAR_GITHUB_REPO_BRANCH = Github Branch for your Repo TF_VAR_GITHUB_AZURE_PIPELINE_PATH = Path in Github Repo for azure-pipelines file. TF_VAR_AZDO_GITHUB_SERVICE_CONNECTION_PAT = Github Personal Token TF_VAR_KUBE_CONFIG_PATH = Path to active kubeconfig file

  Example

  export TF_VAR_AZP_URL=https://dev.azure.com/YourOrg
  export TF_VAR_AZP_TOKEN=.......
  export TF_VAR_AZP_POOL="AzurePipeline"
  export TF_VAR_GITHUB_REPO_NAME="redhat-na-ssa/azure-dev-ops-with-managed-openshift-gitops"
  export TF_VAR_GITHUB_REPO_BRANCH="master"
  export TF_VAR_GITHUB_AZURE_PIPELINE_PATH="/pipeline-example2-terraform/azure/azure-pipelines.yaml"
  export AZDO_PERSONAL_ACCESS_TOKEN=${TF_VAR_AZP_TOKEN}
  export AZDO_ORG_SERVICE_URL=${TF_VAR_AZP_URL}
  export TF_VAR_AZDO_PERSONAL_ACCESS_TOKEN=${TF_VAR_AZP_TOKEN}
  export TF_VAR_AZDO_ORG_SERVICE_URL=${TF_VAR_AZP_URL}
  export TF_VAR_AZDO_GITHUB_SERVICE_CONNECTION_PAT=.........
  export KUBE_CONFIG_PATH=${KUBECONFIG}
  export TF_VAR_KUBE_CONFIG_PATH=${KUBECONFIG}

• Deploy Pipeline Example 2

  cd ./pipeline-example2-terraform/azure
  terraform init --upgrade
  terraform apply -auto-approve
  

• Confirmation

  • Go to the created project in Azure Portal
    

  • Select the create pipeline object

  • Run the Pipeline

  • Check Pipline Status after uccessful Run

  • P.S There is a small error that might show up. "Error: no names or ids specified. Havent yet figured out what causes it. Pipeline should still run successfully.

• The application should also be visible in the ArgoCD Dashboard