17.6 CI/CD, Actions, Jenkins

Q27. Typical pipeline stages?

Lint → test → build → push artifact → deploy staging → (approval) → deploy production.

Q28. Why fail fast?

Cheap checks first so you do not wait for a long image build to discover a typo.

Q29. Why is "CI green" not always "safe for prod"?

Migrations, data shape, cache, traffic patterns, and config still need staging and a rollback plan.

Q30. Where should secrets live in GitHub Actions?

Encrypted Actions secrets / OIDC to cloud roles — never plaintext in YAML.

Q31. Why tag images with git SHA?

Immutable reference: staging and prod can promote the same bits; rollback is redeploy previous SHA.

Q32. When do companies still use Jenkins?

Existing investment, complex self-hosted fleets, or teams standardised on Jenkinsfiles. Actions is common for GitHub-centric repos — know both at concept level.

Q33. Freestyle vs Pipeline job (Jenkins)?

Freestyle is UI-clicky; Pipeline (Jenkinsfile) is code-reviewed automation — prefer Pipeline for repeatability.