What is continuous integration?
Continuous integration is a software development process where developers integrate the new code they've written more frequently throughout the development cycle, adding it to the code base at least once a day. Automated testing is done against each iteration of the build to identify integration issues earlier, when they are easier to fix , which also helps avoid problems at the final merge for the release. Overall, continuous integration helps streamline the build process, resulting in higher-quality software and more predictable delivery schedules.
Continuous integration (CI) vs. continuous delivery (CD) vs. continuous deployment
With continuous integration, each developer integrates their work with the main branch of source code at least once a day (or, preferably, multiple times a day). Continuous delivery, another DevOps practice, instead focuses on delivering any validated changes to the code base—updates, bug fixes, even new features—to users as quickly and safely as possible. Continuous deployment streamlines the process further, using automated testing to validate code base changes, leading to more immediate updates.
Continuous delivery picks up where continuous integration ends, automating the delivery of applications to selected infrastructure environments. It ensures the automation of pushing code changes to different environments, such as development, testing and production.
The following are a few of the most notable benefits that continuous integration provides:
- Continuous and demonstrated progress for improved feedback
- Early and improved error detection and metrics that let you address errors early—sometimes within minutes of check-in
- Improved team collaboration; everyone on the team can change the code, integrate the system and quickly determine conflicts with other parts of the software
- Improved system integration, which reduces surprises at the end of the software development lifecycle
- Fewer parallel changes for merging and testing
- Reduced number of errors during system testing
- Constantly updated systems to test against
CI, agile and DevOps
Agile is a software development practice that improves how software development teams organize themselves, adapt to changes in requirements and release software. Because continuous integration (link resides outside IBM) and agile development (PDF, 153 KB) share many of the same features (e.g., testing automation), it can be useful to talk about continuous integration and agile at the same time. Agile organizes development into smaller groups of work or sprints. When applied in DevOps, these combined practices help to ensure software quality and project flexibility.
Continuous integration requires you to integrate work frequently, often many times per day. You verify integration by an automated build that detects integration errors as early as possible. The build should include run tests as part of verification. Extending the rapid testing to run time tests in an automated testing environment leads naturally towards continuous delivery.
Agile (link resides outside IBM) is also iterative and adapts to change so it can scale and evolve solutions over time. In the context of continuous integration, agile software development is about delivering software iterations based on how you prioritize the value of the features as you continuously integrate.
In the DevOps framework, continuous integration sits at the beginning of the software development process, where you check in your code at least once a day to keep your local copies from drifting too far away from the main branch of the code build. This helps you avoid disastrous merge conflicts that could "break" the build and take the team hours or days to resolve.
Continuous integration serves as a prerequisite for the testing, deployment and release stages of continuous delivery. The entire development team will know within minutes of check-in whether you've created bad code, as the continuous integration service automatically builds and tests your code changes for any errors.
Open source continuous integration tools
The following are some of the most popular continuous integration tools:
- Jenkins: A widely used open source continuous integration tool, Jenkins allows developers to automatically build, integrate and test code as soon as they commit it to the source repository, making it easier for developers to catch bugs early and deploy software faster. The docker plug-in is available on Jenkins.
- Buildbot: Buildbot can automate all aspects of the software development cycle. As a job scheduling system, it queues and executes jobs, and reports results.
- Go: What makes Go stand out from the crowd is the concept of pipelines, which makes the modeling of complex build workflows easy.
- IBM UrbanCode Build
- Travis CI: One of the oldest and most-trusted hosted solutions, it is also available in an on-premises version for the enterprise.
- GitLab CI: An integral part of the open source Rails project, GitLab CI is a free hosted service that provides detailed git repository management with features like access control, issue tracking, code reviews and more.
Conducting continuous integration with open source tools offers many benefits, including the following:
- Hundreds of plug-ins that can support your project
- No cost, giving students, startups and developers working on the side a powerful tool that's easy on the budget
- Customizability, allowing developers to take the foundation of a CI tool and build on top of it to fit the needs
- The ability to modify and redistribute the tools
Open source continuous integration tools to consider for your software development workflow include Jenkins, Go, Buildbot and Travis CI, which you can read about in the next section.
The following hypothetical use case illustrates how two software developers can use continuous integration to improve their DevOps process.
The two developers must communicate with each other on which features work and how. This small team needs regular updates and must be able to integrate and test their code as a whole. Scheduling code check-in and testing take up a lot of development time. An automatic system for continuous integration is needed.
Negotiating when these combinations and tests would happen would eat up a lot of developer time.
To do that they would have to agree on the following:
- When to begin testing code integration
- How to test that the integration has been successful
- How to communicate the results to the team
Continuous integration platforms have default answers to these questions and most allow configuration and setup.
Typically, CI platforms like Jenkins begin integration tests on check-in. When new code is checked in, the CI system will run a set of tests, which can include unit tests and regression tests, and then determine if the code was successfully integrated.
Either way, most CI systems log the integration attempts, success rate and other metrics.
A continuous integration server is a software tool that centralizes all your continuous integration operations and provides a reliable, stable platform for you to build your projects. You can configure and adjust CI servers to build various projects for different platforms. A continuous integration server easily models and visualizes complex workflows (enabling continuous delivery) and provides an intuitive interface for building continuous delivery pipelines. A continuous integration server offers the ability to do the following:
- Run automatic builds, tests and releases in a single place
- Deploy any version, anytime
- Keep configuration orderly
- Support plug-ins to enhance functionality
- Monitor your project’s repository
- Pull changes and perform tasks you defined for successful commit
- Send feedback to the relevant project members with details of the build
The importance of testing
Continuous testing begins when you produce a continuous integration build and a package (also known as an installable entity or packaged entity). It stops when that packaged entity goes into production. Every step from end to end involves a testing suite.
At a minimum, when you only have one test stage, 30 percent of continuous integration involves testing. In reality, continuous integration activities are comprised of 50 percent to 70 percent testing. It used to be that you had to complete testing manually. Now you can use automated tests—the key to successful continuous integration.
As part of automating testing for continuous integration, test-driven development iteratively builds code and tests one use case at a time to ensure test coverage, improve code quality and set the groundwork for continuous delivery. Automated testing tells you whether new code failed one or more of the tests developed across all functional areas of the application. A best practice requires developers to run all or a subset of tests in their local environments, which ensures that developers only commit source code to version control after the new code changes pass their tests. Experience shows that effective regression testing can help avoid unwelcome surprises later.
Continuous integration pipeline
A continuous integration pipeline automates stages of a project's pipeline—such as builds, tests and deployments—in a repeatable way, with minimal human intervention. An automated continuous integration pipeline is essential to streamline the development, testing and deployment of your applications by enabling controls, checkpoints and speed.
The continuous integration process is a critical component of DevOps, which helps you unify your development and operations teams into a shared repository for coding, testing, deploying and supporting software. The following are some CI best practices that can help you succeed:
- Maintain a single source code repository: Use source control management to track and control all the files for building a product. This consolidated code base makes distribution and visibility easier.
- Automate the build: This involves compiling, linking and other processes that produce the build artifacts. Self-testing should also be automated.
- Use daily mainline commits: Force developers to commit their changes to the main development stream at least once every day. Each developer must verify that their working copy is consistent with the main development stream.
- Test in a clone of the production environment: Make the test environment as similar as possible to your final production environment.
- Automate deployment: Implement multiple environments (development, integration, production) to run builds and tests.
Continuous integration and IBM Cloud®
The IBM approach is to define and automate projects and configure security using templates. When a change is made to a library, dependent applications are rebuilt and connected, linked or tied back. Understanding application dependencies will help you modernize your applications.
As organizations accelerate such digital transformation, the need for automation widens across business and IT operations. A move toward greater automation should start with small, measurably successful projects, which you can then scale and optimize for other processes and in other parts of your organization.
Working with IBM, you’ll have access to AI-powered automation capabilities, including prebuilt workflows, to help accelerate innovation by making every process more intelligent.
Take the next step:
— Start managing your software builds, including scaling and configuration, with help from IBM® UrbanCode® Build.
— Read about the five “must-have’s” for automation success (link resides outside IBM) in this HFS Research report.
Get started with an IBM Cloud account today.