liveBook · Manning

CI/CD server

As explained in Automating API Delivery

A build server is a platform that automates tasks in a software delivery pipeline and coordinates its components.

A CI/CD server, also called a build server, runs the tasks in a software delivery pipeline and brings its components together. During continuous integration, it can turn code into software artifacts and run tests after changes are integrated. For API teams, the same platform can also check API definitions and guide documentation before proposed changes are merged. section 8.0 section 8.3

A pull-request (PR) build is a build-server job triggered when someone proposes a change. It runs checks before reviewers spend time on the proposal, then displays its status on the PR. If it fails, the submitter can follow the link to the job results and inspect the logs. The checks and failure thresholds should reflect the team's merge criteria. section 8.3 section 8.7

For an API design proposal, a PR build can:

  • Check the PR title with a script and use a PR template to prompt for the reason behind the change. section 8.3
  • Lint API definitions with Spectral, failing the build on ERROR-level problems; lint Markdown guide documentation according to its configured failure level. section 8.3 section 8.7
  • Compare old and new API definitions for breaking changes with Tufin/oasdiff, failing the build when it detects one. section 8.3 section 8.7

A passing build clears a path for review; it does not replace review. Reviewers still need to examine the build report for non-failing warnings. A code-hosting service's code-owners feature can separately require approval from an API governance team member before affected files are merged. GitHub Actions is one possible build-server platform, not a required pipeline design; some hosting and build-server combinations may need additional PR-build setup. section 8.0 section 8.3 section 8.7

How Automating API Delivery covers it

How deep it goes
Explains PR-build triggers, checks, status, and failure investigation, without detailing a complete server configuration.
Examples
Uses an API team's PR title script, Spectral linting, Tufin/oasdiff checks, and GitHub Actions.
What it assumes
Assumes some familiarity with build servers while briefly defining continuous integration and the deployment pipeline.
What you can do after
Operate/configure it to report PR-check results, fail builds at chosen thresholds, and route remaining issues to reviewers.
How it connects
Connects continuous integration, pull requests, API linting, breaking-change detection, and code-owner approval.

Also explained in

Compare how 4 books explain CI/CD server →

Related concepts

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