AZ-400 - Design and Implement Build and Release Pipelines (53% of the exam) - Section 3.2

Design and implement a dependency and artifact versioning strategy, including semantic versioning (SemVer) and date-based (CalVer).

Distinguish semantic versioning (SemVer) from date-based versioning (CalVer) and apply each to pipeline artifact versioning based on release cadence and consumer expectations. Understand package promotion across feed views so that only versions meeting quality gates reach downstream consumers.

semantic versioningCalVerdependency versioningpipeline artifact versioningpackage promotion

Practice question for this objective

Free sampleDesign and Implement Build and Release Pipelinesmedium

A NuGet package is published to an "Azure Artifacts" feed that has separate views named Local, Prerelease, and Release. After the package passes integration testing, the team wants downstream consumers configured to the Release view to start resolving that exact build without rebuilding or re-uploading it. Which action achieves this?

  • ARebuild the package with the same version number and push it again while targeting the Release view directly, so the binary lands where Release consumers resolve it.
  • BPromote the existing package version to the Release view, so consumers pointed at that view resolve the already-published build with no rebuild or re-upload. Correct
  • CRaise the package version by one patch segment and republish it, so the higher number naturally supersedes the tested build for Release consumers.
  • DDelete the package from the Prerelease view and re-upload the same file into the Release view, so only Release consumers can then resolve the build.
Use feed view promotion to expose an already-published package version to a release audience without rebuilding it. Azure Artifacts feed views are filtered windows over the immutable packages already in a feed. Promoting a version into the Release view changes only its visibility to consumers configured for that view, so the validated build is exposed as is, with no rebuild, re-upload, or version change required.

Why A is wrong: Rebuilding and re-pushing the same version duplicates work the scenario rules out and risks an immutability conflict, since a published version cannot simply be overwritten in the feed.

Why B is correct: Feed views expose a slice of immutable published versions, and promoting a version into the Release view makes Release consumers resolve that exact tested build without any rebuild or re-upload.

Why C is wrong: Bumping the version produces a different, untested build rather than promoting the exact build that passed integration testing, so it changes what consumers receive instead of promoting the validated artifact.

Why D is wrong: Deleting and re-uploading is destructive, loses history, and is exactly the re-upload the scenario forbids, so it is the wrong mechanism even though it ends with the build in Release.

See more AZ-400 practice questions, answers explained.

Exam traps in Design and Implement Build and Release Pipelines

Answers that look right on this material and are not. Each one is a distractor from a different question in the AZ-400 bank for this domain.

  • Adopt strict semantic versioning such as 4.2.0, so each artifact version conveys the level of backward compatibility relative to the previously published build.

    Why it is wrong: SemVer is built to signal compatibility through major, minor, and patch segments, but this team explicitly does not want a compatibility contract, so SemVer answers a question they are not asking.

  • Convert the pre-deployment gate into a manual pre-deployment approval so a release manager re-runs the query by hand each time.

    Why it is wrong: A manual approval is tempting because it also pauses promotion, but it replaces automated re-evaluation with human clicking and does not poll the telemetry, so it fails the unattended re-checking requirement.

  • Add a schedules block to the Release pipeline timed shortly after the CI pipeline usually finishes, because a timed run will pick up the latest CI artifacts once they have been published.

    Why it is wrong: A schedule guesses at timing rather than reacting to the CI completion, so it can start before artifacts exist or long after they are ready, failing the requirement to start as soon as CI completes.

Examworthy is not affiliated with or endorsed by Microsoft. Original, blueprint-aligned practice material only.