Day 17: Deployment Strategies - Hands-On Lab Guide

Lab 1: runOnce Deployment (Baseline) ⚡

Objective: Implement simple runOnce deployment
Time: 25 minutes

Exercise 1.1: Basic runOnce Pipeline

1 Navigate to Project

cd $env:USERPROFILE\Desktop\testing-demo\calculator
notepad azure-pipelines-runonce.yml

2 Create runOnce Pipeline

# runOnce Strategy - Simplest Deployment

trigger:
  - main

stages:
- stage: Build
  jobs:
  - job: BuildJob
    pool:
      vmImage: 'windows-latest'
    steps:
    - task: Maven@3
      inputs:
        goals: 'clean package -DskipTests'
    
    - task: PublishPipelineArtifact@1
      inputs:
        targetPath: 'target'
        artifactName: 'app'

- stage: Deploy
  dependsOn: Build
  jobs:
  - deployment: DeployApp
    displayName: 'runOnce Deployment'
    environment: 'Production'
    pool:
      vmImage: 'windows-latest'
    
    strategy:
      runOnce:
        deploy:
          steps:
          - script: echo "Stop old version"
          - task: DownloadPipelineArtifact@2
          - script: echo "Deploy new version"
          - script: echo "Start application"

Save, commit, push to Azure Repos

✅ Checkpoint: Baseline runOnce deployment working

Lab 2: Rolling Deployment 🔄

Objective: Implement rolling deployment with lifecycle hooks
Time: 40 minutes

Exercise 2.1: Configure Rolling Strategy

1 Create Rolling Pipeline

notepad azure-pipelines-rolling.yml

Add rolling deployment:

# Rolling Deployment Strategy

trigger:
  - main

stages:
- stage: Build
  jobs:
  - job: BuildJob
    pool:
      vmImage: 'windows-latest'
    steps:
    - task: Maven@3
      inputs:
        goals: 'clean package -DskipTests'
    
    - task: PublishPipelineArtifact@1
      inputs:
        targetPath: 'target'
        artifactName: 'app'

- stage: Deploy
  dependsOn: Build
  jobs:
  - deployment: RollingDeploy
    displayName: 'Rolling Deployment'
    environment: 'Production'
    pool:
      vmImage: 'windows-latest'
    
    strategy:
      rolling:
        maxParallel: 2  # Update 2 servers at a time
        
        # Pre-deployment validation
        preDeploy:
          steps:
          - script: |
              echo "================================"
              echo "PRE-DEPLOYMENT CHECKS"
              echo "Checking server health..."
              echo "Backing up configuration..."
              echo "Validation passed!"
              echo "================================"
            displayName: 'Pre-Deploy Health Check'
        
        # Actual deployment
        deploy:
          steps:
          - task: DownloadPipelineArtifact@2
            displayName: 'Download Artifact'
            inputs:
              artifactName: 'app'
              targetPath: '$(Pipeline.Workspace)/app'
          
          - script: |
              echo "================================"
              echo "DEPLOYING TO SERVER"
              echo "Stopping application..."
              timeout /t 2 /nobreak
              echo "Updating files..."
              dir $(Pipeline.Workspace)\app
              timeout /t 2 /nobreak
              echo "Starting application..."
              timeout /t 2 /nobreak
              echo "Server updated successfully!"
              echo "================================"
            displayName: 'Deploy Application'
        
        # Post-deployment validation
        postDeploy:
          steps:
          - script: |
              echo "================================"
              echo "POST-DEPLOYMENT VALIDATION"
              echo "Testing application endpoint..."
              echo "Health check: PASSED"
              echo "Server is healthy and serving traffic"
              echo "================================"
            displayName: 'Post-Deploy Health Check'
        
        # Rollback on failure
        on:
          failure:
            steps:
            - script: |
                echo "DEPLOYMENT FAILED!"
                echo "Rolling back to previous version..."
                echo "Rollback complete"
              displayName: 'Rollback on Failure'
          
          success:
            steps:
            - script: echo "Rolling deployment completed successfully!"
              displayName: 'Success Notification'

Save and close

2 Commit and Run

git add azure-pipelines-rolling.yml
git commit -m "Add rolling deployment strategy"
git push

3 Create Pipeline and Watch Execution

Create new pipeline in Azure DevOps
Select the rolling YAML file
Run and watch lifecycle hooks execute:
✅ preDeploy (health check)
✅ deploy (update application)
✅ postDeploy (validation)
✅ success (completion)

✅ Checkpoint: Rolling deployment with lifecycle hooks working

Lab 3: Canary Deployment 🐤

Objective: Implement gradual user percentage rollout
Time: 40 minutes

Exercise 3.1: Configure Canary Strategy

1 Create Canary Pipeline

notepad azure-pipelines-canary.yml

# Canary Deployment Strategy

trigger:
  - main

stages:
- stage: Build
  jobs:
  - job: BuildJob
    pool:
      vmImage: 'windows-latest'
    steps:
    - task: Maven@3
      inputs:
        goals: 'clean package -DskipTests'
    
    - task: PublishPipelineArtifact@1
      inputs:
        targetPath: 'target'
        artifactName: 'app'

- stage: DeployCanary
  displayName: 'Canary Deployment to Production'
  dependsOn: Build
  
  jobs:
  - deployment: CanaryDeploy
    displayName: 'Gradual Rollout'
    environment: 'Production'
    pool:
      vmImage: 'windows-latest'
    
    strategy:
      canary:
        increments: [10, 25, 50, 100]  # 10% → 25% → 50% → 100%
        
        preDeploy:
          steps:
          - script: |
              echo "========================================"
              echo "CANARY DEPLOYMENT - STARTING"
              echo "Target: $(strategy.increment)%% of users"
              echo "========================================"
            displayName: 'Canary Start'
        
        deploy:
          steps:
          - task: DownloadPipelineArtifact@2
            inputs:
              artifactName: 'app'
          
          - script: |
              echo "========================================"
              echo "ROUTING $(strategy.increment)%% TRAFFIC"
              echo "Deploying new version..."
              echo "Configuring load balancer..."
              echo "New version live for $(strategy.increment)%% of users"
              echo "========================================"
            displayName: 'Deploy to $(strategy.increment)%'
        
        postDeploy:
          steps:
          - script: |
              echo "========================================"
              echo "MONITORING CANARY: $(strategy.increment)%%"
              echo "Checking error rates..."
              echo "Monitoring response times..."
              echo "Analyzing user feedback..."
              timeout /t 3 /nobreak
              echo "Metrics look good! ✅"
              echo "Ready for next increment"
              echo "========================================"
            displayName: 'Monitor $(strategy.increment)%'
        
        on:
          failure:
            steps:
            - script: |
                echo "❌ CANARY FAILED at $(strategy.increment)%%"
                echo "Rolling back traffic to stable version..."
                echo "Deployment stopped!"
              displayName: 'Canary Rollback'
          
          success:
            steps:
            - script: |
                echo "🎉 CANARY DEPLOYMENT COMPLETE!"
                echo "100%% of users on new version"
              displayName: 'Canary Success'

Save and close

2 Push and Create Pipeline

git add azure-pipelines-canary.yml
git commit -m "Add canary deployment strategy"
git push

3 Watch Canary Rollout

Pipeline will execute in phases:

Phase 1: Deploy to 10% of users
Monitor and validate
Phase 2: Increase to 25%
Monitor and validate
Phase 3: Increase to 50%
Monitor and validate
Phase 4: Complete rollout to 100%

Note: In this demo, we simulate the incremental rollout. In production, you'd configure load balancer routing rules.

✅ Checkpoint: Canary deployment with incremental rollout

Lab 4: Blue-Green Deployment Concept 🔵🟢

Objective: Understand blue-green deployment workflow
Time: 35 minutes

Exercise 4.1: Simulated Blue-Green Pipeline

1 Create Blue-Green Pipeline

notepad azure-pipelines-bluegreen.yml

# Blue-Green Deployment (Simulated)

trigger:
  - main

stages:
- stage: Build
  jobs:
  - job: BuildJob
    pool:
      vmImage: 'windows-latest'
    steps:
    - task: Maven@3
      inputs:
        goals: 'clean package -DskipTests'
    
    - task: PublishPipelineArtifact@1
      inputs:
        targetPath: 'target'
        artifactName: 'app'

# Deploy to Green (inactive environment)
- stage: DeployGreen
  displayName: 'Deploy to Green Environment'
  dependsOn: Build
  
  jobs:
  - deployment: DeployToGreen
    environment: 'Green-Environment'
    pool:
      vmImage: 'windows-latest'
    
    strategy:
      runOnce:
        deploy:
          steps:
          - task: DownloadPipelineArtifact@2
            inputs:
              artifactName: 'app'
          
          - script: |
              echo "========================================"
              echo "DEPLOYING TO GREEN ENVIRONMENT"
              echo "Blue is still serving all traffic"
              echo "Green is being updated..."
              echo "Deployment to Green complete!"
              echo "Green is ready but not receiving traffic"
              echo "========================================"
            displayName: 'Deploy to Green'
          
          - script: |
              echo "VALIDATING GREEN ENVIRONMENT"
              echo "Running smoke tests..."
              echo "Testing database connectivity..."
              echo "Testing API endpoints..."
              echo "All tests passed! ✅"
            displayName: 'Validate Green'

# Switch traffic from Blue to Green
- stage: SwitchTraffic
  displayName: 'Switch Traffic to Green'
  dependsOn: DeployGreen
  
  jobs:
  - deployment: TrafficSwitch
    environment: 'Production'
    pool:
      vmImage: 'windows-latest'
    
    strategy:
      runOnce:
        deploy:
          steps:
          - script: |
              echo "========================================"
              echo "SWITCHING TRAFFIC"
              echo "Current: Blue (v1.0) - 100%% traffic"
              echo "Target: Green (v2.0) - 0%% traffic"
              echo ""
              echo "Updating load balancer..."
              timeout /t 2 /nobreak
              echo ""
              echo "Traffic switched! ⚡"
              echo "Blue (v1.0) - 0%% traffic (idle)"
              echo "Green (v2.0) - 100%% traffic ✅"
              echo "========================================"
            displayName: 'Switch Load Balancer'
          
          - script: |
              echo "MONITORING NEW VERSION"
              echo "Watching error rates..."
              echo "Monitoring response times..."
              echo "System healthy! ✅"
            displayName: 'Monitor After Switch'

Save and close

2 Create Green Environment

Go to Pipelines → Environments
Create new environment: Green-Environment

3 Run Blue-Green Pipeline

git add azure-pipelines-bluegreen.yml
git commit -m "Add blue-green deployment"
git push

Create pipeline and watch three-phase execution:

Build: Create artifact
Deploy to Green: Update inactive environment
Switch Traffic: Make Green active

✅ Checkpoint: Blue-green workflow demonstrated

Lab 5: Compare Deployment Strategies 📊

Objective: Run and compare all three strategies
Time: 30 minutes

Exercise 5.1: Compare Execution

1 Run All Three Pipelines

Run azure-pipelines-runonce.yml
Run azure-pipelines-rolling.yml
Run azure-pipelines-canary.yml

2 Compare Characteristics

Aspect	runOnce	Rolling	Canary
Stages	1 deploy stage	1 with hooks	Multiple phases
Time	~30 seconds	~2 minutes	~5 minutes
Safety	Basic	Good	Excellent
Complexity	Simple	Medium	Complex

✅ Checkpoint: Understand tradeoffs between strategies

🎯 Lab Summary

What You've Accomplished:

✅ Implemented runOnce deployment (baseline)
✅ Created rolling deployment with lifecycle hooks
✅ Built canary deployment with gradual rollout
✅ Simulated blue-green traffic switching
✅ Compared all deployment strategies
✅ Understood tradeoffs for each approach

Key Concepts Mastered:

Rolling: Update servers incrementally
Blue-Green: Two environments with instant switch
Canary: Gradual user percentage rollout
Lifecycle Hooks: preDeploy, deploy, postDeploy
Failure Handling: Automatic rollback on errors

🎭 Day 17: Deployment Strategies - Hands-On Guide

Lab 1: runOnce Deployment (Baseline) ⚡

Exercise 1.1: Basic runOnce Pipeline

Lab 2: Rolling Deployment 🔄

Exercise 2.1: Configure Rolling Strategy

Lab 3: Canary Deployment 🐤

Exercise 3.1: Configure Canary Strategy

Lab 4: Blue-Green Deployment Concept 🔵🟢

Exercise 4.1: Simulated Blue-Green Pipeline

Lab 5: Compare Deployment Strategies 📊

Exercise 5.1: Compare Execution

🎯 Lab Summary

What You've Accomplished:

Key Concepts Mastered:

🎉 Day 17 Complete!