top » management » project management » agile » continuous delivery » continuous integration

13 Principles of Continuous Integration

John Spacey, updated on August 22, 2018

Continuous integration is the practice of merging the code of multiple software developers several times a day. This is a foundation for agile development and continuous delivery whereby code is incrementally developed, built, deployed and tested such that it is regularly available for release. The following are basic principles of continuous integration.

Short Development Cycles

Organize work into short development cycles that produce working functionality.

Version Control

Use of a version control tool that provides versioning and a history of builds and files.

Everyone Commits Daily

In principle, all developers commit working code to a shared baseline each day.

Every Commit is Built & Tested

Every commit to the baseline triggers a build. Developers unit test their code in the common build.

Fix Broken Builds Immediately

Fixing broken builds is prioritized as there are more code changes coming at a continual pace and things get messy if build problems pile up.

Rapidly Detect Problems

The entire philosophy behind continuous integration is that problems should be detected as early as possible. This principle can be applied in countless ways. For example, automated regression tests may be organized such that commonly broken functions are tested first.

Automate the Build

The build is automated and designed for speed.

Automate Testing

Toil is reduced by automating testing where possible including build testing, unit testing and automated acceptance testing that includes regression tests.

Automate Deployment

Environments can be deployed with the press of a button. In some cases, deployment is automatically triggered by criteria such as a green light in automated testing.

Use a Deployment Pipeline

Testing environments progress from those that are completely unstable and changed several times a day to environments such as user acceptance testing and production staging that are deployed less often to provide a stable environment for testing. The following is a typical deployment pipeline process.

Test Reflects Production

Every effort is made to have test environments such as UAT reflect production, often at reduced scale. This includes elements such as platforms, systems, infrastructure, architecture and data.

Builds are Transparent

Everyone can see what a build contains including a list of implemented requirements and/or fixed defects.

Builds are Available

Everyone can access a build. For example, a user acceptance testing team can access an environment with the latest build at any time.

Overview: Continuous Integration
Type	Continuous Delivery
Definition	The practice of merging the code of multiple software developers several times a day.
Related Concepts	Continuous Delivery » Agile » Incrementalism » Toil » Regression Tests » Deployment Pipeline »