Topic 7 of 12
CI/CD pipelines
Learn how a pipeline turns every commit into a tested, packaged and deployable change, using GitHub Actions to see each moving part.
After this topic you can: Write a GitHub Actions workflow that builds, tests and deploys a change, and find out why a pipeline failed or never started.
8 lessons · 68 min · 0 of 8 done
Best read after: Git and collaboration, Containers with Docker, Python and scripting for ops
Draft. This lesson is written and waiting for a person to run every command and review it. It is safe to learn from, but check the official documentation for anything you depend on.
Lessons, in order
- 1The anatomy of a pipelineCoreUnderstand the five parts every CI/CD pipeline has (trigger, jobs, steps, runners, artifacts) by first writing one as a plain shell script.7 min
- 2Writing a GitHub Actions workflowCoreWrite a workflow file with events, jobs, steps and needs, push it to a throwaway repository and see it run in the Actions tab.10 min
- 3Build, test and artifactsCoreSplit a pipeline into build and test jobs, pass the built files between them as an artifact, and see why you test what you ship.9 min
- 4Caching and faster pipelinesCoreCut pipeline time by caching dependencies with a content-based key, and cancel runs that a newer push has made pointless.8 min
- 5Secrets and variables in pipelinesCoreKeep tokens out of git and logs: store them as secrets, scope them to environments, see how masking works and fails, and meet OIDC.9 min
- 6Environments and approvalsAdvancedModel staging and production as named environments, so a deploy job can pause for a human approval and only then receive its secrets.7 min
- 7Deployment strategiesAdvancedCompare rolling, blue-green and canary releases, what each costs, how rollback works, and practise a blue-green switch locally.8 min
- 8Debugging failed pipelinesCoreFind why a run failed or never started: read the first red step, re-run with debug logging, and check filters and file location.10 min