Interview Q&A

What are pipeline templates and how are they used?

Answer:

Templates let you reuse pipeline steps across projects — great for standardization.

Example:

You can create a reusable file:

📄 .azure-pipelines/build-template.yml

steps:

• script: dotnet restore
• script: dotnet build

Then in your main pipeline:

extends:

template: .azure-pipelines/build-template.yml

This keeps your YAML DRY (Don’t Repeat Yourself) and consistent.

Answer: "ApplicationInsights": { "ConnectionString": "InstrumentationKey=xxxx-xxxx-xxxx" }

What interviewers expect

• A clear definition tied to DevOps in Azure DevOps projects
• Trade-offs (performance, maintainability, security, cost)
• When you would and would not use it in production

Real-world example

In a production Azure DevOps application, teams apply this when handling user-facing features or integration boundaries. For example, you might use it during a sprint where reliability and observability matter—logging metrics, validating edge cases, and documenting the decision in an ADR so future developers understand why the approach was chosen.

How to explain in the interview

1. Define the concept in one or two sentences.
2. Context — where it fits in Azure DevOps architecture.
3. Example — a specific project, bug, or performance win.
4. Trade-off — what you gain vs what you sacrifice.

Tip: Practice aloud on Toolliyo mock interview or the Interview Q&A section before your real interview.

Connection → Azure Resource Manager

What interviewers expect

• A clear definition tied to DevOps in Azure DevOps projects
• Trade-offs (performance, maintainability, security, cost)
• When you would and would not use it in production

Real-world example

In a production Azure DevOps application, teams apply this when handling user-facing features or integration boundaries. For example, you might use it during a sprint where reliability and observability matter—logging metrics, validating edge cases, and documenting the decision in an ADR so future developers understand why the approach was chosen.

How to explain in the interview

1. Define the concept in one or two sentences.
2. Context — where it fits in Azure DevOps architecture.
3. Example — a specific project, bug, or performance win.
4. Trade-off — what you gain vs what you sacrifice.

Tip: Practice aloud on Toolliyo mock interview or the Interview Q&A section before your real interview.

dotnet restore

What interviewers expect

• A clear definition tied to DevOps in Azure DevOps projects
• Trade-offs (performance, maintainability, security, cost)
• When you would and would not use it in production

Real-world example

In a production Azure DevOps application, teams apply this when handling user-facing features or integration boundaries. For example, you might use it during a sprint where reliability and observability matter—logging metrics, validating edge cases, and documenting the decision in an ADR so future developers understand why the approach was chosen.

How to explain in the interview

1. Define the concept in one or two sentences.
2. Context — where it fits in Azure DevOps architecture.
3. Example — a specific project, bug, or performance win.
4. Trade-off — what you gain vs what you sacrifice.

Tip: Practice aloud on Toolliyo mock interview or the Interview Q&A section before your real interview.

In your pipeline or locally, publish the .nupkg file using:

dotnet nuget push "MyLibrary.1.0.0.nupkg" --source "MyFeed"

• -api-key az

or use the Azure Pipelines task:

• task: NuGetCommand@2

inputs:

command: 'push'

Follow:

packagesToPush: '$(Build.ArtifactStagingDirectory)/*.nupkg'

publishVstsFeed: 'MyProject/MyFeed'

dd SonarQube tasks to your build pipeline:

• task: SonarQubePrepare@5

inputs:

SonarQube: 'MySonarServiceConnection'

scannerMode: 'MSBuild'

projectKey: 'MyProject'

projectName: 'MyApp'

• script: dotnet build
• task: SonarQubeAnalyze@5
• task: SonarQubePublish@5

inputs:

pollingTimeoutSec: '300'

Answer: <<<<<<< HEAD old code ======= new code >>>>>>> feature/login

What interviewers expect

• A clear definition tied to DevOps in Azure DevOps projects
• Trade-offs (performance, maintainability, security, cost)
• When you would and would not use it in production

Real-world example

In a production Azure DevOps application, teams apply this when handling user-facing features or integration boundaries. For example, you might use it during a sprint where reliability and observability matter—logging metrics, validating edge cases, and documenting the decision in an ADR so future developers understand why the approach was chosen.

How to explain in the interview

1. Define the concept in one or two sentences.
2. Context — where it fits in Azure DevOps architecture.
3. Example — a specific project, bug, or performance win.
4. Trade-off — what you gain vs what you sacrifice.

Tip: Practice aloud on Toolliyo mock interview or the Interview Q&A section before your real interview.

How do you handle secrets in pipelines (Azure Key Vault integration)?

Answer:

You never store passwords directly in YAML — instead, use Azure Key Vault or variable

groups linked to Key Vault.

Example:

• Create a Key Vault in Azure.
• Add secrets like SqlPassword.
• In pipeline, add Key Vault as a variable group.

YAML:

variables:

• group: MyKeyVaultVariables

Secrets are pulled securely during build and never exposed in logs.

nother team adds the feed to their project’s NuGet.config and installs the package like

ny normal NuGet dependency.

3⃣ How do you version your NuGet packages in a CI/CD pipeline?

Versioning is usually handled automatically in the build pipeline — so every build produces

unique version number.

You can version packages using:

• Build ID, Git commit hash, or semantic versioning (e.g., 1.2.3).

Example (YAML):

variables:

versionMajor: 1

versionMinor: 0

versionPatch: $(Build.BuildId)

steps:

• script: dotnet pack MyLibrary.csproj -c Release
• p:PackageVersion=$(versionMajor).$(versionMinor).$(versionPatch)

displayName: 'Create NuGet package'

Then publish it:

• task: NuGetCommand@2

inputs:

command: 'push'

packagesToPush: '**/*.nupkg'

publishVstsFeed: 'MyProject/MyFeed'

Example scenario:

Each time your CI pipeline runs, it generates a package version like 1.0.45 or 1.0.46 —

ensuring consistent versioning and avoiding overwriting old packages.

4⃣ How do you manage dependencies between multiple .NET projects

using Azure Artifacts?

When you have multiple .NET projects that depend on each other, Azure Artifacts acts as

your internal package registry.

Typical approach:

Quality gates are automated checks that your code must pass before it can be deployed.

Typical gates:

• Code must pass all tests.
• Code coverage ≥ 80%.
• No critical SonarQube issues.
• Build must succeed.

Ways to enforce them:

• Branch policies: Block PRs until build and tests pass.
• SonarQube gate: Fail the pipeline if quality gate fails.
• Pipeline conditions: Only deploy if all checks succeed.

Example (YAML):

• task: SonarQubePublish@5

inputs:

pollingTimeoutSec: '300'

condition: succeeded()

Example scenario:

If SonarQube reports a “Failed Quality Gate” due to high code duplication, the deployment

to QA is blocked until issues are fixed.

5⃣ How do you perform automated smoke testing after deployment?

Smoke tests verify that your deployed app is running and key endpoints work — without

doing deep functional testing.

You can run these as post-deployment steps in your release pipeline.

Example (PowerShell):

• task: PowerShell@2

inputs:

targetType: 'inline'

script: |

$response = Invoke-WebRequest

if ($response.StatusCode -ne 200) {

throw "Smoke test failed!"

}

Example scenario:

fter deploying your API to QA, the pipeline runs a smoke test that checks:

• /health endpoint returns 200
• Database connection works

If it fails, the pipeline stops and alerts the team.

✅ Pro Tip:

Combine all testing and quality practices like this:

CommonLibrary in its .csproj.

What interviewers expect

• A clear definition tied to DevOps in Azure DevOps projects
• Trade-offs (performance, maintainability, security, cost)
• When you would and would not use it in production

Real-world example

In a production Azure DevOps application, teams apply this when handling user-facing features or integration boundaries. For example, you might use it during a sprint where reliability and observability matter—logging metrics, validating edge cases, and documenting the decision in an ADR so future developers understand why the approach was chosen.

How to explain in the interview

1. Define the concept in one or two sentences.
2. Context — where it fits in Azure DevOps architecture.
3. Example — a specific project, bug, or performance win.
4. Trade-off — what you gain vs what you sacrifice.

Tip: Practice aloud on Toolliyo mock interview or the Interview Q&A section before your real interview.

Versioning is usually handled automatically in the build pipeline — so every build produces

a unique version number.

You can version packages using:

• Build ID, Git commit hash, or semantic versioning (e.g., 1.2.3).

Example (YAML):

variables:

versionMajor: 1

versionMinor: 0

versionPatch: $(Build.BuildId)

Follow:

steps:

• script: dotnet pack MyLibrary.csproj -c Release
• p:PackageVersion=$(versionMajor).$(versionMinor).$(versionPatch)

displayName: 'Create NuGet package'

Then publish it:

• task: NuGetCommand@2

inputs:

command: 'push'

packagesToPush: '**/*.nupkg'

publishVstsFeed: 'MyProject/MyFeed'

Example scenario:

Each time your CI pipeline runs, it generates a package version like 1.0.45 or 1.0.46 —

ensuring consistent versioning and avoiding overwriting old packages.

4⃣ How do you manage dependencies between multiple .NET projects

using Azure Artifacts?

When you have multiple .NET projects that depend on each other, Azure Artifacts acts as

your internal package registry.

Typical approach:

Answer: dd your feed’s URL to NuGet.config: <add key="MyFeed" value=" ndex.json" />

What interviewers expect

• A clear definition tied to DevOps in Azure DevOps projects
• Trade-offs (performance, maintainability, security, cost)
• When you would and would not use it in production

Real-world example

In a production Azure DevOps application, teams apply this when handling user-facing features or integration boundaries. For example, you might use it during a sprint where reliability and observability matter—logging metrics, validating edge cases, and documenting the decision in an ADR so future developers understand why the approach was chosen.

How to explain in the interview

1. Define the concept in one or two sentences.
2. Context — where it fits in Azure DevOps architecture.
3. Example — a specific project, bug, or performance win.
4. Trade-off — what you gain vs what you sacrifice.

Tip: Practice aloud on Toolliyo mock interview or the Interview Q&A section before your real interview.

Quality gates are automated checks that your code must pass before it can be deployed.

Typical gates:

• Code must pass all tests.
• Code coverage ≥ 80%.
• No critical SonarQube issues.
• Build must succeed.

Ways to enforce them:

• Branch policies: Block PRs until build and tests pass.
• SonarQube gate: Fail the pipeline if quality gate fails.
• Pipeline conditions: Only deploy if all checks succeed.

Example (YAML):

• task: SonarQubePublish@5

inputs:

pollingTimeoutSec: '300'

condition: succeeded()

Example scenario:

If SonarQube reports a “Failed Quality Gate” due to high code duplication, the deployment

to QA is blocked until issues are fixed.

5⃣ How do you perform automated smoke testing after deployment?

Smoke tests verify that your deployed app is running and key endpoints work — without

doing deep functional testing.

You can run these as post-deployment steps in your release pipeline.

Follow:

Example (PowerShell):

• task: PowerShell@2

inputs:

targetType: 'inline'

script: |

$response = Invoke-WebRequest

if ($response.StatusCode -ne 200) {

throw "Smoke test failed!"

Example scenario:

After deploying your API to QA, the pipeline runs a smoke test that checks:

• /health endpoint returns 200
• Database connection works

If it fails, the pipeline stops and alerts the team.

✅ Pro Tip:

Combine all testing and quality practices like this:

How do you trigger a build automatically after code check-in?

Answer:

Add a trigger section in your YAML file, or use the GUI option “Enable continuous

integration.”

Example (YAML):

trigger:

branches:

include:

• main
• develop

Every push to main or develop automatically triggers a build.

You can also trigger manually or from a pull request.

What are Azure DevOps Services vs Azure DevOps Server?

Answer:

• Azure DevOps Services → Cloud version (hosted by Microsoft). You don’t worry

about servers or upgrades.

• Azure DevOps Server → On-premises version (you manage the infrastructure).

Example:

If you’re a bank that must keep data on-premises, you’d use Azure DevOps Server.
If you’re a startup that wants zero maintenance, Azure DevOps Services is ideal.

zureResourceManagerConnection: 'MyServiceConnection'

resourceGroupName: 'MyRG'

location: 'East US'

csmFile: 'infra/main.json'

Example scenario:

Your IaC pipeline uses the “ProdServiceConnection” to deploy Bicep templates to the

production resource group, without developers ever seeing credentials.

4⃣ How can you automate environment creation and teardown?

You can automate creating and destroying environments (like dev, test, or staging) using

pipeline logic and IaC tools.

✅ Example (Bicep – Create Environment):

• task: AzureCLI@2

inputs:

zureSubscription: 'MyServiceConnection'

scriptType: 'bash'

inlineScript: |

z group create --name MyRG --location eastus

z deployment group create --resource-group MyRG

• -template-file infra/main.bicep

✅ Example (Terraform – Teardown Environment):

• task: TerraformCLI@1

inputs:

command: 'destroy'

workingDirectory: 'infra'

commandOptions: '-auto-approve'

environmentServiceName: 'MyServiceConnection'

Example scenario:

Each pull request automatically spins up a temporary Azure environment (App Service +

Database) for testing.

Once the PR is closed or merged, the pipeline runs a teardown job to delete the resources

— saving cost and keeping Azure clean.

✅ Pro Tip:

For professional IaC pipelines in Azure DevOps:

• Use Bicep for Azure-native infrastructure.
• Use Terraform for multi-cloud or complex setups.
• Always deploy with a service connection using least-privilege roles.
• Automate environment lifecycle (create → test → destroy).

Security & Compliance Advanced /

Real-world Scenarios

1⃣ How do you secure secrets, keys, and connection strings in Azure

Pipelines?

You should never hardcode secrets (like passwords, connection strings, or API keys) in

pipeline YAML files or scripts.

Instead, you store them securely using Azure DevOps Variable Groups or Azure Key

Vault integration.

Ways to secure secrets:

• Use secret variables in pipelines:

→ In Azure DevOps → Pipelines → Variables → Mark variable as secret.

• Use Azure Key Vault to fetch secrets securely at runtime.
• Use pipeline permissions to control who can view or edit variables.

Example (YAML):

variables:

• group: 'ProdSecrets'

steps:

• script: echo "Connecting to DB..."

env:

ConnectionString: $(DB_Connection)

Example scenario:

You’re deploying an app that needs a SQL connection string.

You store that in Azure Key Vault and reference it securely in the pipeline — so it never

ppears in logs or code.

2⃣ What is Azure Key Vault and how is it used with pipelines?

Azure Key Vault is a secure store for secrets, keys, and certificates.

zure DevOps can connect to it to retrieve secrets dynamically during builds or

deployments.

Steps to use it:

zureSubscription: 'MyServiceConnection'

scriptType: 'bash'

scriptLocation: 'inlineScript'

inlineScript: |

echo "Checking App Insights availability..."

z monitor metrics list --resource myapp --metric

requests/count

Example scenario:

fter a new deployment, your team monitors App Insights dashboards — noticing that

response times increased by 20%.

They can quickly roll back or optimize performance before users notice.

✅ Pro Tip:

Combine these tools for full visibility:

• Azure Pipelines logs → Diagnose build/deployment failures.
• Azure Monitor + App Insights → Track live app performance.
• Power BI / Analytics → Measure DevOps effectiveness.
• Dashboards → Share live health status with your whole team.

Infrastructure as Code (IaC) &

utomation

1⃣ How do you provision infrastructure in Azure using ARM, Bicep, or

Terraform via Azure DevOps?

You can automate infrastructure deployment directly from Azure Pipelines using ARM

templates, Bicep files, or Terraform scripts.

Each option defines your infrastructure as code — meaning servers, networks, and

resources are described in files and deployed consistently through pipelines.

✅ Example (using Bicep in YAML):

trigger:

• main

pool:

vmImage: 'ubuntu-latest'

steps:

• task: AzureCLI@2

inputs:

zureSubscription: 'MyServiceConnection'

scriptType: 'bash'

scriptLocation: 'inlineScript'

inlineScript: |

z deployment group create \

• -resource-group MyRG \
• -template-file infrastructure/main.bicep \
• -parameters environment=dev appName=myapp

✅ Example (using Terraform):

• task: TerraformInstaller@1

inputs:

terraformVersion: '1.7.0'

• task: TerraformCLI@1

inputs:

command: 'init'

workingDirectory: 'infra'

• task: TerraformCLI@1

inputs:

command: 'apply'

workingDirectory: 'infra'

commandOptions: '-auto-approve'

environmentServiceName: 'MyServiceConnection'

Example scenario:

DevOps pipeline runs when a pull request is merged into main, automatically provisioning

n Azure App Service, Storage Account, and SQL Database using Bicep.

2⃣ What is the difference between ARM templates and Bicep?

Feature ARM Templates Bicep

Language JSON Domain-specific (simpler syntax)

Readability Complex, verbose Clean and concise

Reusability Harder (manual

nesting)

Supports modules easily

Tooling Native in Azure Compiles into ARM JSON

Learning curve Steep Easier for beginners

In short:

Bicep is the modern, simplified language for ARM templates.

It makes infrastructure code shorter and easier to read — but still deploys through the same

zure Resource Manager (ARM) engine.

Example comparison:

RM Template (JSON):

{

"resources": [

{

"type": "Microsoft.Storage/storageAccounts",

"name": "[parameters('storageName')]",

"location": "[resourceGroup().location]",

"sku": { "name": "Standard_LRS" },

"kind": "StorageV2"

}

}

Bicep (same thing):

resource storageAccount

'Microsoft.Storage/storageAccounts@2022-09-01' = {

name: storageName

location: resourceGroup().location

sku: { name: 'Standard_LRS' }

kind: 'StorageV2'

}

Example scenario:

cloud engineer switches from ARM to Bicep and cuts a 300-line template down to just 80

lines — making it much easier to maintain in Git.

3⃣ How do you use Azure Service Connections for IaC deployments?

An Azure Service Connection securely stores credentials that pipelines use to connect to

your Azure subscription.

It’s like giving your pipeline “keys” to deploy resources in Azure — without hardcoding

credentials.

How to create one:

rtifact links, etc.).

Example scenario:

If your .NET build failed during dotnet test, you open the “Test” step log — Azure

DevOps shows which specific test failed, the error message, and even a link to the exact line

of code if integrated with Repos.

Tip:

You can also use log filters (e.g., “Errors only”) to focus on failed tasks quickly.

2⃣ How do you troubleshoot failed builds or deployments?

Troubleshooting starts by identifying where and why the pipeline failed.

Steps:

zure DevOps?

You can deploy Infrastructure as Code (IaC) directly from your pipeline using ARM or

Bicep files.

Example (YAML):

• task: AzureResourceManagerTemplateDeployment@3

inputs:

deploymentScope: 'Resource Group'

zureResourceManagerConnection: 'MyServiceConnection'

subscriptionId: 'xxxx-xxxx-xxxx'

ction: 'Create Or Update Resource Group'

resourceGroupName: 'my-rg'

location: 'East US'

templateLocation: 'Linked artifact'

csmFile: 'infrastructure/main.bicep'

overrideParameters: '-appName myapp -sku S1'

Example scenario:

Before deploying your .NET app, your pipeline provisions a new App Service, SQL

Database, and Storage Account automatically.

9⃣ How can you use Azure CLI or PowerShell tasks in your release

pipeline?

You can run custom scripts using AzureCLI@2 or PowerShell@2 tasks to automate

dvanced tasks.

Example (Azure CLI):

• task: AzureCLI@2

inputs:

zureSubscription: 'MyServiceConnection'

scriptType: 'bash'

scriptLocation: 'inlineScript'

inlineScript: |

z webapp restart --name my-webapp --resource-group my-rg

Example (PowerShell):

• task: PowerShell@2

inputs:

targetType: 'inline'

script: |

Write-Host "Performing custom cleanup..."

Remove-Item -Path $(Build.ArtifactStagingDirectory)\temp

• Recurse -Force

Example scenario:

fter deploying your app, you might use a PowerShell script to clear old log files or an Azure

CLI command to restart the web app.

✅ Pro Tip:

solid Release Pipeline usually includes:

• Staged deployments (Dev → QA → Prod)
• Environment-specific configs
• Pre-deployment approvals
• Slot-based zero-downtime deployment
• Rollback and infrastructure provisioning

Testing & Quality

1⃣ How do you integrate unit tests, integration tests, or UI tests in Azure

Pipelines?

You can run all types of automated tests (unit, integration, UI) inside your build or release

pipelines using tasks like DotNetCoreCLI@2, Visual Studio Test@2, or custom

scripts.

Example (Unit Tests):

• task: DotNetCoreCLI@2

inputs:

command: 'test'

projects: '**/*UnitTests.csproj'

publishTestResults: true

Example (Integration Tests):

• Typically run after the app is deployed to a test environment.
• You can use a task like:
• script: dotnet test

./Tests/IntegrationTests/IntegrationTests.csproj

Example (UI Tests):

• Tools like Selenium or Playwright can be integrated as post-deployment steps in

your pipeline:

• script: npx playwright test

Real-life scenario:

fter building a .NET API, the pipeline runs unit tests.

Then, in a release pipeline, once the app is deployed to the “QA” environment, Selenium

tests verify that the login screen works.

2⃣ What is code coverage, and how do you view it in Azure DevOps?

Code coverage measures how much of your code is executed during tests — it helps you

see which parts of your application aren’t tested.

Example (for .NET):

• script: dotnet test --collect:"XPlat Code Coverage"

This generates a coverage report (coverage.cobertura.xml), which Azure DevOps can

display in the Test Results → Code Coverage tab.

Example scenario:

Your pipeline might show that 82% of your code is covered by tests.

You can then aim to increase that by adding more unit tests for uncovered modules.

3⃣ How do you use SonarQube or other tools for static code analysis?

SonarQube checks your code for bugs, code smells, and security vulnerabilities.

Integration steps:

zureSubscription: 'MyServiceConnection'

ppName: 'my-webapp'

deployToSlotOrASE: true

resourceGroupName: 'my-rg'

slotName: 'staging'

Then use:

• task: AzureAppServiceManage@0

inputs:

ction: 'Swap Slots'

SourceSlot: 'staging'

ResourceGroupName: 'my-rg'

WebAppName: 'my-webapp'

7⃣ How do you perform rollback in case of a failed deployment?

There are several ways:

.NET Core app gets built, zipped, and deployed automatically to the Staging slot of an

zure Web App after successful testing.

5⃣ How do you manage configuration transformations (e.g., web.config or

ppsettings.json)?

You can use XML or JSON transformation tasks in your release pipeline to adjust settings

per environment.

Example (Classic):

• Add a File Transform or Replace Tokens task.
• Replace values like database connection strings or API URLs for QA, UAT, or

Production.

Example (YAML):

• task: FileTransform@2

inputs:

folderPath: '$(System.DefaultWorkingDirectory)/drop'

xmlTransformation: true

jsonTargetFiles: '**/appsettings*.json'

Example scenario:

In appsettings.Production.json, you might replace the connection string with your

production database credentials automatically during deployment.

6⃣ How do you handle deployment slots in Azure App Service?

Azure App Service supports deployment slots — like Staging and Production.

Best practice:

You can automate creating and destroying environments (like dev, test, or staging) using

pipeline logic and IaC tools.

✅ Example (Bicep – Create Environment):

• task: AzureCLI@2

inputs:

azureSubscription: 'MyServiceConnection'

scriptType: 'bash'

inlineScript: |

az group create --name MyRG --location eastus

az deployment group create --resource-group MyRG

• -template-file infra/main.bicep

✅ Example (Terraform – Teardown Environment):

• task: TerraformCLI@1

Follow:

inputs:

command: 'destroy'

workingDirectory: 'infra'

commandOptions: '-auto-approve'

environmentServiceName: 'MyServiceConnection'

Example scenario:

Each pull request automatically spins up a temporary Azure environment (App Service +

Database) for testing.

Once the PR is closed or merged, the pipeline runs a teardown job to delete the resources

— saving cost and keeping Azure clean.

✅ Pro Tip:

For professional IaC pipelines in Azure DevOps:

• Use Bicep for Azure-native infrastructure.
• Use Terraform for multi-cloud or complex setups.
• Always deploy with a service connection using least-privilege roles.
• Automate environment lifecycle (create → test → destroy).

Security & Compliance Advanced /

Real-world Scenarios

1⃣ How do you secure secrets, keys, and connection strings in Azure

Pipelines?

You should never hardcode secrets (like passwords, connection strings, or API keys) in

pipeline YAML files or scripts.

Instead, you store them securely using Azure DevOps Variable Groups or Azure Key

Vault integration.

Ways to secure secrets:

Follow:

• Use secret variables in pipelines:

→ In Azure DevOps → Pipelines → Variables → Mark variable as secret.

• Use Azure Key Vault to fetch secrets securely at runtime.
• Use pipeline permissions to control who can view or edit variables.

Example (YAML):

variables:

• group: 'ProdSecrets'

steps:

• script: echo "Connecting to DB..."

env:

ConnectionString: $(DB_Connection)

Example scenario:

You’re deploying an app that needs a SQL connection string.

You store that in Azure Key Vault and reference it securely in the pipeline — so it never

appears in logs or code.

2⃣ What is Azure Key Vault and how is it used with pipelines?

Azure Key Vault is a secure store for secrets, keys, and certificates.

Azure DevOps can connect to it to retrieve secrets dynamically during builds or

deployments.

Steps to use it:

You can automate infrastructure deployment directly from Azure Pipelines using ARM

templates, Bicep files, or Terraform scripts.

Each option defines your infrastructure as code — meaning servers, networks, and

resources are described in files and deployed consistently through pipelines.

✅ Example (using Bicep in YAML):

trigger:

• main

pool:

vmImage: 'ubuntu-latest'

steps:

• task: AzureCLI@2

inputs:

azureSubscription: 'MyServiceConnection'

scriptType: 'bash'

scriptLocation: 'inlineScript'

inlineScript: |

az deployment group create \

• -resource-group MyRG \
• -template-file infrastructure/main.bicep \
• -parameters environment=dev appName=myapp

✅ Example (using Terraform):

• task: TerraformInstaller@1

inputs:

terraformVersion: '1.7.0'

Follow:

• task: TerraformCLI@1

inputs:

command: 'init'

workingDirectory: 'infra'

• task: TerraformCLI@1

inputs:

command: 'apply'

workingDirectory: 'infra'

commandOptions: '-auto-approve'

environmentServiceName: 'MyServiceConnection'

Example scenario:

A DevOps pipeline runs when a pull request is merged into main, automatically provisioning

an Azure App Service, Storage Account, and SQL Database using Bicep.

2⃣ What is the difference between ARM templates and Bicep?

Feature ARM Templates Bicep

Language JSON Domain-specific (simpler syntax)

Readability Complex, verbose Clean and concise

Reusability Harder (manual

nesting)

Supports modules easily

Tooling Native in Azure Compiles into ARM JSON

Learning curve Steep Easier for beginners

In short:

Bicep is the modern, simplified language for ARM templates.

It makes infrastructure code shorter and easier to read — but still deploys through the same

Azure Resource Manager (ARM) engine.

Example comparison:

Follow:

ARM Template (JSON):

"resources": [

"type": "Microsoft.Storage/storageAccounts",

"name": "[parameters('storageName')]",

"location": "[resourceGroup().location]",

"sku": { "name": "Standard_LRS" },

"kind": "StorageV2"

Bicep (same thing):

resource storageAccount

'Microsoft.Storage/storageAccounts@2022-09-01' = {

name: storageName

location: resourceGroup().location

sku: { name: 'Standard_LRS' }

kind: 'StorageV2'

Example scenario:

A cloud engineer switches from ARM to Bicep and cuts a 300-line template down to just 80

lines — making it much easier to maintain in Git.

3⃣ How do you use Azure Service Connections for IaC deployments?

An Azure Service Connection securely stores credentials that pipelines use to connect to

your Azure subscription.

It’s like giving your pipeline “keys” to deploy resources in Azure — without hardcoding

credentials.

How to create one:

Follow:

Troubleshooting starts by identifying where and why the pipeline failed. Steps:

What interviewers expect

• A clear definition tied to DevOps in Azure DevOps projects
• Trade-offs (performance, maintainability, security, cost)
• When you would and would not use it in production

Real-world example

In a production Azure DevOps application, teams apply this when handling user-facing features or integration boundaries. For example, you might use it during a sprint where reliability and observability matter—logging metrics, validating edge cases, and documenting the decision in an ADR so future developers understand why the approach was chosen.

How to explain in the interview

1. Define the concept in one or two sentences.
2. Context — where it fits in Azure DevOps architecture.
3. Example — a specific project, bug, or performance win.
4. Trade-off — what you gain vs what you sacrifice.

Tip: Practice aloud on Toolliyo mock interview or the Interview Q&A section before your real interview.