Implementing CI/CD Pipelines: Concepts, Scenarios, and Project Practices

The article introduces the author's perspective on CI/CD and presents multiple practical scenarios for applying continuous integration and continuous deployment within projects. It emphasizes the importance of automating build, test, and deployment processes to reduce manual effort and improve software quality.

CI/CD Basic Concepts

CI (Continuous Integration) refers to automatically running tests and builds after each code push, ensuring that new changes do not break the main branch. CD can mean either Continuous Delivery—automatically preparing code for release—or Continuous Deployment—automatically releasing validated code to production environments.

CICD Scenarios

3.1 Compilation & Deployment

Automates the build and deployment steps after code submission, replacing manual operations. Benefits include eliminating waiting time between build and deployment and providing real‑time notifications.

3.2 Unit Testing

Uses JUnit to write backend unit tests, generating test reports and JaCoCo coverage metrics. Benefits include higher test coverage and clear insight into code exercised by tests.

3.3 Code Scanning

Integrates SonarQube and the internal EOS platform to automatically scan code quality on each commit, producing reports and tracking issues until they are resolved.

3.4 Automated Testing

Leverages Python‑based frameworks such as EasyOne, DeepTest, and JMeter to run automated regression tests, collect reports, and send notifications.

3.5 Full‑Link Testing

Combines the above scenarios into an end‑to‑end pipeline: code commit triggers compilation, code scanning, unit tests, deployment, followed by automated tests and result notifications.

Project Practice

The author describes using JD's internal Bamboo platform as the CI/CD pipeline backbone, detailing how pipelines are built for compilation‑deployment, unit testing, and automated testing. Example Maven configuration snippets are provided:

<plugins>
    <plugin>
        <groupId>org.apache.maven.plugins</groupId>
        <artifactId>maven-surefire-plugin</artifactId>
        <version>2.5</version>
    </plugin>
</plugins>

<plugin>
    <groupId>org.apache.maven.plugins</groupId>
    <artifactId>maven-surefire-plugin</artifactId>
    <version>2.5</version>
    <configuration>
        <skipTests>false</skipTests>
    </configuration>
</plugin>

mvn clean test -Dmaven.test.failure.ignore=true

<plugin>
    <groupId>org.jacoco</groupId>
    <artifactId>jacoco-maven-plugin</artifactId>
    <version>0.8.2</version>
    <executions>
        <execution>
            <goals>
                <goal>prepare-agent</goal>
            </goals>
        </execution>
        <execution>
            <id>report</id>
            <phase>test</phase>
            <goals>
                <goal>report</goal>
            </goals>
        </execution>
    </executions>
</plugin>

These configurations enable Maven to run unit tests, generate Surefire reports, and produce JaCoCo coverage data.

The article also outlines deployment triggers (manual, commit‑based, scheduled, or chained pipelines) and reports on the adoption results, such as thousands of pipeline executions and a 60% increase in unit test effectiveness.

Future Planning

Plans include expanding automated testing with DeepTest, tightening quality gates in code scanning, and further integrating CI/CD pipelines across more projects.