AZ-400 Practice Test

Answer choices

1. A. Add more callers without changing the template
2. B. Use a deployment job targeting the environment, keep reusable structure in the template, and supply secrets through protected runtime values
3. C. Convert each environment to a trigger
4. D. Print the parameter after expansion so masking can find it

Wrong-answer review

• A. Add more callers without changing the template: More callers multiply both defects rather than correcting them.
• C. Convert each environment to a trigger: Triggers decide when runs start and do not represent deployment targets.
• D. Print the parameter after expansion so masking can find it: Printing sensitive data increases exposure and cannot retroactively make it safe.

Answer choices

1. A. A rolling deployment that replaces all virtual-machine instances in one batch.
2. B. A canary deployment with increments, postRouteTraffic health checks, and an on: failure hook.
3. C. A runOnce deployment with no post-deployment hook.
4. D. A continuous-integration trigger that gates the rollout by branch name.

Answer choices

1. A. Repository write permission automatically grants pipeline and Azure access
2. B. Check repository permission, pipeline resource authorization, service connection permission, and Azure RBAC as separate layers
3. C. Make the repository public to grant Azure access
4. D. Add a Git tag because tags authorize deployments

Answer choices

1. A. Use every commit message as authoritative release scope
2. B. Replace the wiki with more chat channels
3. C. Write notes manually from memory after each incident
4. D. Generate notes from governed work and version evidence, publish durable documentation, and have chat notifications link to it

Answer choices

1. A. Instrument both components and propagate correlation context through the queue so Application Insights can join their telemetry
2. B. Monitor only queue length
3. C. Assign the same static operation ID to every request
4. D. Use deployment history as distributed tracing

Answer choices

1. A. License analysis only
2. B. A source tag that marks the image reviewed
3. C. CodeQL rerun after production deployment
4. D. Pre-promotion image and secret scanning with severity policy, credential revocation, and governed exceptions

Answer choices

1. A. LFS deletes all historical large blobs as soon as tracking is enabled
2. B. Branch protection converts prior commits to pointers
3. C. Future tracked commits use pointer files and external LFS storage, while old history needs a deliberate coordinated migration
4. D. Release tags move the large content out of Git

Answer choices

1. A. Reward the frequency increase because one metric defines DevOps success
2. B. Pair the metric with outcome and reliability measures and use separate queries for delivery, quality, security, and operations decisions
3. C. Combine all signals into an undocumented success score
4. D. Stop measuring outcomes and count commits instead

Answer choices

1. A. The average proves users are unaffected
2. B. Top-level duration is enough and dependency spans add no value
3. C. Scale CPU because every latency problem is compute
4. D. Tail users are affected, and the dependency span localizes the dominant delay for targeted investigation

Answer choices

1. A. Grant release managers and developers force push on main
2. B. Use only repository read permission because it defines merge checks
3. C. Put reviewers in a variable group
4. D. Use branch permissions for contribution and narrow bypass authority, plus branch policies for reviewer and build gates