CI/CD Interview Questions

Run fast, reliable checks on every pull request (lint, unit, component, API and a small UI smoke set), a broader regression after merge against a deployed environment, and slow or expensive suites (full E2E, cross-browser, performance, security scans) on a schedule or before release.

Build and static checks, unit and component tests, contract tests, image build and scan, deploy to a test environment, API and integration tests, then promotion with smoke tests and canary checks. Each gate is an objective, automated rule, such as new-code coverage or zero critical vulnerabilities.

Trigger on pull_request and push, use a matrix to split the suite into shards, cache dependencies through setup-node or setup-java, set fail-fast false so every shard reports, and upload reports and traces with actions/upload-artifact@v4 under if: always() and a unique name per shard.

Detect flakiness with data (pass-on-retry, history across runs), quarantine flaky tests out of the blocking set with an owner and deadline, fix root causes such as timing, shared data and environment instability, and treat retries as a detection tool rather than a fix. Track the flake rate as a team metric.