Trunk-based development

As explained in Platform Engineering on Kubernetes

Trunk-based development is a source-control practice in which the main branch remains continuously releasable through small, validated changes and short-lived feature branches.

Trunk-based development keeps the main branch continuously releasable. Developers use short-lived feature branches, validate changes before merging, and avoid introducing changes that break the main branch’s build or release process. section 3.3

A trunk-based workflow connects branch validation to release automation:

• Run a pipeline for each feature branch or pull request to verify that the changes build, test, and can be released against the main branch. section 3.4.1
• Run the service pipeline after changes merge to main, producing artifacts from the latest code base. section 3.4.1
• Treat a failing main-branch pipeline as an urgent problem; teams should focus on restoring the releasable state instead of merging more breakage. section 3.4.1
• Create a Git tag for a release so the resulting artifact can be traced to the specific main-branch commit that produced it. section 3.4.1 section 3.4

A merge-to-release flow

A developer completes and tests a feature on a feature branch, opens a pull request, and waits for its validation pipeline. After review and a successful check, the change merges into main. The main-branch pipeline builds and tests the service, publishes binary artifacts, builds the container image, and can produce a tagged release. Each successful merge therefore supplies a tested, traceable release candidate. section 3.4.1 section 3.4

Figure 3.4 Service pipelines for main branch and feature branches

Use the pipeline diagram to compare validation on feature branches with release-producing execution on the main branch. section 3.4.1

How Platform Engineering on Kubernetes covers it

How deep it goes

Defines branch and pipeline behavior through release tagging and validation, but does not prescribe one universal pipeline implementation.

Examples

Uses pull-request checks, main-branch pipelines, Git tags, binary artifacts, container images, unit tests, coverage checks, and security scans.

What it assumes

Assumes a Git repository and automated service pipeline, while defining branch validation and release-readiness practices.

What you can do after

Operate a Git workflow that validates feature changes, protects main, and produces traceable releases after successful merges.

How it connects

Connects feature branches, pull requests, continuous integration, service pipelines, release tags, artifact repositories, and container images.

liveBookmanning.com

Read Platform Engineering on Kubernetes on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Infrastructure as Code, Patterns and Practices: With examples in Python and TerraformTrunk-based development is a workflow that integrates small, incremental changes directly into a shared main branch with short-lived or no feature branches.
• 

  liveBookmanning.com
  Effective Platform EngineeringTrunk-based development is a version-control practice in which developers integrate small changes frequently into a shared mainline.
• 

  liveBookmanning.com
  The AI Pocket BookA development workflow where changes are integrated directly into the main branch with supporting practices to avoid breaking production.

Compare how 4 books explain trunk-based development →

Related concepts

• branchuses · A branch is an independently evolving line of development in a version-control repository, typically represented by a reference to a sequence of changes or commits.
• deployment pipelineuses · A sequence of automated or manual stages that builds, tests or validates, packages, and delivers software or models toward deployment.
• feature-based developmentalternative to · Feature-based development is a software and infrastructure workflow that isolates changes on separate branches before review and merging.
• feature flagconfigures · A feature flag is a configuration control that conditionally enables, disables, or selects a software feature or behavior.
• driftaddresses · Drift is a change over time in a system's behavior, state, or characteristics relative to a reference, intended condition, or expected baseline.

This explainer was generated by Manning's knowledge graph from the text of Platform Engineering on Kubernetes. The section links point to the parts of the book it draws on.