Create a Multi-branch Pipeline with GitLab

    In KubeSphere 3.1.x and later, you can create a multi-branch pipeline with GitLab in your DevOps project. This tutorial demonstrates how to create a multi-branch pipeline with GitLab.

    • You need to have a GitLab account and a account.
    • You need to create a workspace, a DevOps project and a user (). This user must be invited to the DevOps project with the operator role. For more information, refer to Create Workspaces, Projects, Users and Roles.

    Hands-on Lab

    1. Log in to the KubeSphere console as project-regular. Go to your DevOps project and create the following credentials in Credentials under DevOps Project Settings. For more information about how to create credentials, see Credential Management.

      Note

      If there are any special characters such as @ and $ in your account or password, they can cause errors as a pipeline runs because they may not be recognized. In this case, you need to encode your account or password on some third-party websites first, such as . After that, copy and paste the output for your credential information.

    2. After creation, you can see the credentials in the list.

    Step 2: Modify the Jenkinsfile in your GitLab repository

    1. Log in to GitLab and create a public project. Click Import project/repository, select Repo by URL to enter the URL of , select Public for Visibility Level, and then click Create project.

    2. In the project just created, create a new branch from the master branch and name it gitlab-demo.

    3. In the gitlab-demo branch, click the file Jenkinsfile-online in the root directory.

    4. Click Edit, change GITHUB_CREDENTIAL_ID, GITHUB_ACCOUNT, and @github.com to , GITLAB_ACCOUNT, and @gitlab.com respectively, and then edit the following items. You also need to change the value of branch in the push latest and deploy to dev stages to gitlab-demo.

      ItemValueDescription
      GITLAB_CREDENTIAL_IDgitlab-idThe Name you set in KubeSphere for your GitLab account. It is used to push tags to your GitLab repository.
      DOCKERHUB_NAMESPACEfelixnooReplace it with your Docker Hub’s account name. It can be the Organization name under the account.
      GITLAB_ACCOUNTfelixnooReplace it with your GitLab account name. It can also be the account’s Group name.

      Note

    5. Click Commit changes to update this file.

    You need to create two projects, such as kubesphere-sample-dev and kubesphere-sample-prod, which represent the development environment and the production environment respectively. For more information, refer to Create a Pipeline Using a Jenkinsfile.

    Step 4: Create a pipeline

    1. Log in to the KubeSphere web console as project-regular. Go to your DevOps project and click Create to create a new pipeline.

    2. Provide the basic information in the displayed dialog box. Name it gitlab-multi-branch and select a code repository.

    3. On the Advanced Settings tab, scroll down to Script Path. Change it to Jenkinsfile-online and then click Create.

      Note

      The field specifies the Jenkinsfile path in the code repository. It indicates the repository’s root directory. If the file location changes, the script path also needs to be changed.

    1. Click Run on the right. In the displayed dialog box, select gitlab-demo from the drop-down list and add a tag number such as v0.0.2. Click OK to trigger a new run.

      Note

      The pipeline pauses at the stage deploy to dev. You need to click Proceed manually. Note that the pipeline will be reviewed three times as deploy to dev, push with tag, and deploy to production are defined in the Jenkinsfile respectively.

    Step 6: Check the pipeline status

    1. In the Task Status tab, you can see how a pipeline is running. Check the pipeline running logs by clicking View Logs in the upper-right corner.

    2. You can see the dynamic log output of the pipeline, including any errors that may stop the pipeline from running. For each stage, you can click it to inspect logs, which can also be downloaded to your local machine for further analysis.

    1. The Docker image built through the pipeline has been successfully pushed to Docker Hub, as it is defined in the Jenkinsfile. In Docker Hub, you will find the image with the tag v0.0.2 that is specified before the pipeline runs.

    2. At the same time, a new tag is generated in GitLab.