Updated on 2024-11-18 GMT+08:00

Step 8: Configuring a Pipeline for Continuous Delivery

CodeArts Pipeline provides a visualized and customizable software pipeline for automatic delivery. It supports multiple task types, such as code check, build, and deployment tasks.

As the project progresses, each stage (build, release, and deployment) becomes more and more standardized. However, each stage is relatively independent and incomplete and cannot deliver business value directly. Only by effectively connecting each stage to form a complete continuous delivery pipeline can we truly improve the efficiency and quality of software release and continuously create business value.

This section describes how Chris, a developer, connects code check, build, and deployment for continuous delivery.

Introduction to Preset Pipelines

There are five pipeline tasks preset in the sample project. You can view and use them as needed.

Table 1 Preset pipeline tasks

Preset Pipeline Task

Description

phoenix-workflow

A basic pipeline task

phoenix-workflow-test

Pipeline task corresponding to the test environment

phoenix-workflow-work

Pipeline task corresponding to the Worker function

phoenix-workflow-result

Pipeline task corresponding to the Result function

phoenix-workflow-vote

Pipeline task corresponding to the Vote function

Configuring and Executing a Pipeline

A pipeline usually consists of multiple stages. You can add multiple jobs to each stage.

  1. Configure a pipeline.

    1. Go to the Phoenix Mall project and choose CICD > Pipeline.
    2. Find pipeline phoenix-workflow. Click and choose Edit.
    3. Add a code check stage.
      1. Click between Code Source and Build to add a stage.
      2. Click next to Stage_1. In the Edit Stage window, enter the stage name Check and click Confirm.
        Figure 1 Editing the stage name
      3. Click Job.

        In the New Job window, click Add next to the Check extension.

      4. Select the phoenix-codecheck-worker task and click OK.

        The check task has three modes. This procedure uses the default mode Full. You can change the mode as required.

        • Full: All files in the code repository are scanned.
        • Incremental (last commit): Incremental check is performed based on the latest commit file.
        • Incremental (last success): Incremental check is performed based on the changed files since the latest access control was passed.
    4. Configure a deployment task.

      Click the deployment task name, select the associated build task phoenix-sample-ci, and check the values of configuration items.

      • The configurations of task phoenix-sample-standalone must be the same as those on the Parameters page of the task with the same name in CodeArts Deploy.
      • The configurations of task phoenix-cd-cce must be the same as those on the Parameters page of the task with the same name in CodeArts Deploy.

      Two deployment tasks are added in this example. If you selected only one deployment mode in preceding steps, keep the corresponding task and delete the other one.

    5. Click Save.

  2. Go to the CCE console if you have configured deployment task phoenix-cd-cce in 1. Locate the target cluster and click its name to go to the Overview page.

    Choose Workloads from the navigation pane, click the Deployments tab, and verify that no record exists in the list.

    If there are records in the list, select all records, click Delete, select all options, and click Yes to clear the records in the list.

  3. Return to the pipeline list page. Click in the row where phoenix-workflow is located, and click Execute in the window that is displayed to start the pipeline.

    If is displayed, the task is successfully executed.

    If the task fails to be executed, check the failure cause in the failed task. You can open the step details page to view the task logs and rectify the faults based on the logs.

Configuring Pass Conditions

To control the code quality, the code must be scanned and the number of errors must be within a reasonable range before being released. By adding quality gates, you can effectively automate the control process.

  1. On the details page of pipeline task phoenix-workflow, click Edit.
  2. In the Check stage, click Pass Conditions.
  3. In the Pass Conditions dialog box, click Add next to Pass-Conditions-of-Standard-Policies.
  4. Select SystemPolicy and click OK.
  5. Click Save and Execute.

    If the number of check issues does not meet the pass condition, the pipeline task fails to be executed.

Configuring Code Changes to Automatically Trigger a Pipeline

Through the following configuration, code changes can automatically trigger pipeline execution, implementing continuous project delivery.

  1. On the details page of pipeline task phoenix-workflow, click Edit.
  2. Click the Execution Plan tab, select Code commit, select master from the Filter Branch drop-down list box, and click Save.
  3. Modify the code and push it to the master branch to check whether the pipeline task is automatically executed.