AZ-400 practice questions

Question 1. You need to trace a production change back to its requirement, code review, and tests. What enables this?

1. End-to-end traceability linking work items, commits, PRs, builds, and releases
2. Tagging each release with a semantic version number
3. Requiring squash merges on every pull request
4. Enabling verbose logging on the build agents

Question 2. One release spans four repositories, and the team wants a single plan broken into two-week sprints. What should you set up?

1. A milestone in each of the four repositories
2. An organisation-owned GitHub Project with an iteration field
3. A wiki page listing the four repositories and their owners
4. A protected branch in each repository

Question 3. You want to visualise where work is piling up across board columns over time. Which chart helps?

1. A pie chart of languages
2. A pricing estimate
3. A cumulative flow diagram
4. A Gantt of staff holidays

Question 4. You want to embed a rendered flowchart directly inside a project wiki page. Which syntax do you use?

1. Raw HTML canvas
2. A screenshot only
3. A binary .vsdx upload only
4. Mermaid in Markdown

Question 5. In Azure Repos, how do you require two reviewers and a passing build before a PR can complete on main?

1. Configure a branch policy on main
2. Add a resource lock
3. Create a release pipeline gate
4. Use Git LFS

Question 6. A repository is bloated and slow to clone because it stores large binary assets. What should you implement?

1. More reviewers
2. A branch policy
3. Azure Artifacts feed
4. Git Large File Storage (LFS)

Question 7. A regulated product ships four times a year and has to keep patching version 8.2 while version 9.0 is being built. Which branching strategy fits, and what does it cost you?

1. Release branching, at the cost of landing every fix on the release branch and on main
2. Trunk-based development, because a single trunk removes the need to patch older versions at all
3. Feature branching, because every fix already gets its own branch
4. No branching, because tags on main are enough to identify each shipped version

Question 8. A CI job on a very large repository spends most of its run time cloning, and it only ever needs the latest commit. What is the cheapest fix?

1. Move every source file into Git LFS
2. Use a shallow clone so only recent history is fetched
3. Register the repository with Scalar on the build agent
4. Split the repo into dozens of smaller ones

Question 9. Which is GitHub's integrated package hosting service?

1. GitHub Packages
2. Azure Artifacts
3. App Configuration
4. Key Vault

Question 10. A pipeline must use licensed software and reach an on-premises database on a private network. Which agent fits?

1. A self-hosted agent/runner
2. A canary deployment
3. A feed view
4. A Microsoft-hosted agent

Question 11. Which Azure Pipelines construct records deployment history and lets you apply runOnce/rolling/canary strategies against an environment?

1. A variable group
2. A wiki
3. A deployment job
4. A feed view

Question 12. Old pipeline runs and artifacts are consuming storage indefinitely. What should you configure?

1. A shorter default branch history
2. A retention policy for runs and artifacts
3. Pipeline caching for dependencies
4. A self-hosted agent with a larger disk

Question 13. A GitHub Actions job needs to comment on a PR in the same repo without any stored credential. What should it use?

1. A user password
2. The built-in GITHUB_TOKEN
3. A service principal secret
4. A long-lived personal access token

Question 14. You want GitHub Actions to deploy to Azure with no secret stored in the repo at all. Which approach?

1. An SSH key in a variable
2. Workload identity federation (OpenID Connect)
3. A personal access token
4. A shared service account password

Question 15. GitHub detects a vulnerable open-source dependency and can open a PR to bump it. What is this?

1. A canary
2. Defender for Containers
3. CodeQL
4. Dependabot

Question 16. You ship a closed-source product. A scan flags a transitive dependency published under a strong copyleft licence. What is the risk being raised?

1. Distributing the product would oblige you to release your own source under the same licence
2. The package will stop receiving security updates
3. The package cannot be mirrored in an internal feed
4. The licence prohibits commercial use of any kind

Question 17. Where are Azure Monitor logs stored and queried with KQL?

1. A Key Vault
2. A variable group
3. An Azure Artifacts feed
4. A Log Analytics workspace

Question 18. Which Azure Monitor experience gives targeted telemetry for AKS/container workloads?

1. Application Insights
2. Storage Insights
3. VM Insights
4. Container Insights

Question 19. A developer never hears about failed GitHub Actions runs on a repository they contribute to, while a teammate hears about everything on it. Where is that difference set?

1. In the workflow file's concurrency settings
2. In the repository's branch protection rules
3. In each user's GitHub notification settings, which govern watching and whether Actions runs notify them
4. In the organisation's billing settings

Question 20. A request passes through several microservices and you need to find which one adds the latency. What do you use?

1. A canary release
2. A deployment slot
3. A KQL summarize on logins
4. Distributed tracing in Application Insights