AZ-400 Deployment Patterns 2 — Questions and Answers
Question 1: In a ring deployment, what is the primary purpose of 'Ring 0'?
- Deploy to all production users simultaneously
- Deploy to internal Microsoft/team users first for early validation (Correct answer)
- Deploy to a random 10% sample of users
- Deploy only to test environments
Correct answer: Deploy to internal Microsoft/team users first for early validation
Ring 0 targets internal users (dogfooding) so the team validates changes before any external exposure.
Question 2: Which Azure DevOps feature allows you to gate a deployment stage until specific work items are linked to the release?
- Branch policies
- Release gates with work item query (Correct answer)
- Approval policies
- Deployment slots
Correct answer: Release gates with work item query
Release gates can query Azure Boards work items so a stage only proceeds when required items are in the correct state.
Question 3: A team uses feature flags to decouple deployment from release. Which tool is natively integrated with Azure DevOps for feature flag management?
- LaunchDarkly only
- Azure App Configuration with feature management (Correct answer)
- GitHub Actions secrets
- Azure Key Vault only
Correct answer: Azure App Configuration with feature management
Azure App Configuration provides a feature management library that integrates natively with Azure DevOps pipelines and .NET applications.
Question 4: During a canary deployment in AKS, you want to route exactly 20% of traffic to the new version. Which Kubernetes resource controls this traffic split?
- ConfigMap
- Ingress with NGINX annotations or a service mesh weight (Correct answer)
- PersistentVolumeClaim
- ResourceQuota
Correct answer: Ingress with NGINX annotations or a service mesh weight
NGINX Ingress canary annotations or a service mesh (like Istio) VirtualService weights control the percentage of traffic routed to each deployment.
Question 5: What does 'deployment frequency' measure in DORA metrics, and what does a 'Elite' rating require?
- Number of environments; more than 5 environments
- How often code is deployed to production; on-demand or multiple times per day (Correct answer)
- Rollback rate; less than 1% rollbacks
- Team size; more than 20 developers
Correct answer: How often code is deployed to production; on-demand or multiple times per day
Deployment frequency measures how often an organization successfully releases to production; Elite performers deploy on-demand or multiple times per day.
Question 6: You need to perform a blue-green deployment for an Azure App Service. Which App Service feature enables zero-downtime slot swaps?
- Auto-scale rules
- Deployment slots with swap (Correct answer)
- WebJobs
- Application Insights live metrics
Correct answer: Deployment slots with swap
Azure App Service deployment slots let you warm up the new version and then swap slots atomically, achieving zero-downtime blue-green deployments.
Question 7: In a shadow deployment, what happens to traffic sent to the shadow (dark) service?
- It affects real users and replaces live traffic
- Responses are discarded and users receive responses from the live service only (Correct answer)
- It triggers automatic rollback if errors are detected
- It is cached and replayed later
Correct answer: Responses are discarded and users receive responses from the live service only
Shadow deployments mirror real traffic to the new version but discard the shadow's responses so users are unaffected while production behavior is tested.
AZ-400 Deployment Patterns 2
In a ring deployment, what is the primary purpose of 'Ring 0'?