Application Insights Instrumentation

This skill helps you integrate Azure Application Insights into your applications for comprehensive monitoring and telemetry.

When to Use This Skill

Use this skill when you need to:

• Add monitoring and telemetry to ASP.NET Core applications
• Instrument Node.js applications with Application Insights
• Configure custom metrics and logging
• Set up performance tracking and dependency monitoring
• Implement application health monitoring

Capabilities

• Instrument ASP.NET Core applications with Application Insights SDK
• Configure telemetry for Node.js applications
• Set up custom metrics and logging
• Configure connection strings and instrumentation keys
• Enable dependency tracking and request monitoring
• Implement custom telemetry tracking

Prerequisites

• Azure subscription with Application Insights resource
• Target application (ASP.NET Core or Node.js)
• Application hosted in Azure (App Service, Container Apps, etc.) or on-premises
• Understanding of your application's monitoring requirements

Implementation Guide

For ASP.NET Core Applications

Step 1: Add the NuGet Package

dotnet add package Microsoft.ApplicationInsights.AspNetCore

Step 2: Configure in Program.cs

// In Program.cs (ASP.NET Core 6+)
var builder = WebApplication.CreateBuilder(args);

// Add Application Insights telemetry
builder.Services.AddApplicationInsightsTelemetry(options =>
{
    options.ConnectionString = builder.Configuration["ApplicationInsights:ConnectionString"];
});

var app = builder.Build();

Step 3: Add Configuration

Add to appsettings.json:

{
  "ApplicationInsights": {
    "ConnectionString": "InstrumentationKey=your-key;IngestionEndpoint=https://your-region.in.applicationinsights.azure.com/"
  },
  "Logging": {
    "ApplicationInsights": {
      "LogLevel": {
        "Default": "Information",
        "Microsoft": "Warning"
      }
    }
  }
}

Step 4: Use Custom Telemetry

public class MyService
{
    private readonly TelemetryClient _telemetryClient;
    
    public MyService(TelemetryClient telemetryClient)
    {
        _telemetryClient = telemetryClient;
    }
    
    public async Task ProcessOrderAsync(Order order)
    {
        using var operation = _telemetryClient.StartOperation<RequestTelemetry>("ProcessOrder");
        
        try
        {
            // Track custom event
            _telemetryClient.TrackEvent("OrderProcessing", new Dictionary<string, string>
            {
                { "OrderId", order.Id },
                { "CustomerId", order.CustomerId }
            });
            
            // Your business logic here
            await ProcessOrder(order);
            
            // Track custom metric
            _telemetryClient.TrackMetric("OrderValue", order.TotalAmount);
            
            operation.Telemetry.Success = true;
        }
        catch (Exception ex)
        {
            _telemetryClient.TrackException(ex);
            operation.Telemetry.Success = false;
            throw;
        }
    }
}

For Node.js Applications

Step 1: Install Package

npm install applicationinsights

Step 2: Initialize at Startup

// At the very beginning of your app (before other requires)
const appInsights = require('applicationinsights');

appInsights.setup(process.env.APPLICATIONINSIGHTS_CONNECTION_STRING)
    .setAutoDependencyCorrelation(true)
    .setAutoCollectRequests(true)
    .setAutoCollectPerformance(true, true)
    .setAutoCollectExceptions(true)
    .setAutoCollectDependencies(true)
    .setAutoCollectConsole(true)
    .setUseDiskRetryCaching(true)
    .setSendLiveMetrics(true)
    .start();

const client = appInsights.defaultClient;

module.exports = client;

Step 3: Track Custom Telemetry

const appInsights = require('./appInsights');

async function processOrder(order) {
    const startTime = Date.now();
    
    try {
        // Track custom event
        appInsights.trackEvent({
            name: 'OrderProcessing',
            properties: {
                orderId: order.id,
                customerId: order.customerId
            }
        });
        
        // Your business logic
        await handleOrder(order);
        
        // Track metric
        const duration = Date.now() - startTime;
        appInsights.trackMetric({ name: 'OrderProcessingTime', value: duration });
        
        // Track custom metric
        appInsights.trackMetric({ name: 'OrderValue', value: order.totalAmount });
        
    } catch (error) {
        // Track exception
        appInsights.trackException({ exception: error });
        throw error;
    }
}

Best Practices

Security

• Never hardcode connection strings - Use environment variables or Azure Key Vault
• Use Managed Identity when running in Azure for authentication
• Rotate instrumentation keys regularly if not using connection strings
• Filter sensitive data before sending telemetry
• Implement least-privilege access to Application Insights resources

Performance

• Enable sampling in high-volume applications to control costs:

  builder.Services.AddApplicationInsightsTelemetry(options =>
  {
      options.EnableAdaptiveSampling = true;
  });
  

• Use async methods for telemetry operations
• Batch telemetry when tracking multiple events
• Configure appropriate flush intervals

Monitoring

• Set up custom properties for better filtering:

  var telemetryInitializer = new CustomTelemetryInitializer();
  builder.Services.AddSingleton<ITelemetryInitializer>(telemetryInitializer);
  

• Create custom metrics for business KPIs
• Configure alerts in Azure Portal for critical patterns
• Use dependency tracking to identify bottlenecks
• Enable live metrics for real-time monitoring

Cost Management

• Use sampling to reduce ingestion volume (90-95% reduction typical)
• Filter noisy telemetry at source
• Set retention policies appropriate to your needs
• Monitor ingestion rates and costs in Azure Portal

Common Patterns

Pattern 1: Track Business Metrics

public class OrderMetricsTracker
{
    private readonly TelemetryClient _telemetry;
    
    public void TrackOrderCompleted(Order order)
    {
        _telemetry.TrackEvent("OrderCompleted", new Dictionary<string, string>
        {
            { "OrderId", order.Id },
            { "PaymentMethod", order.PaymentMethod },
            { "ShippingCountry", order.ShippingAddress.Country }
        });
        
        _telemetry.TrackMetric("OrderRevenue", order.TotalAmount);
        _telemetry.TrackMetric("OrderItems", order.Items.Count);
    }
}

Pattern 2: Track User Actions

function trackUserAction(action, details) {
    appInsights.trackEvent({
        name: `User_${action}`,
        properties: {
            userId: details.userId,
            action: action,
            page: details.page,
            timestamp: new Date().toISOString()
        }
    });
}

Pattern 3: Performance Monitoring

public async Task<Result> PerformOperationAsync()
{
    using var operation = _telemetryClient.StartOperation<DependencyTelemetry>("DatabaseQuery");
    operation.Telemetry.Type = "SQL";
    operation.Telemetry.Target = "ProductionDB";
    
    try
    {
        var result = await _database.QueryAsync();
        operation.Telemetry.Success = true;
        return result;
    }
    catch (Exception ex)
    {
        operation.Telemetry.Success = false;
        _telemetryClient.TrackException(ex);
        throw;
    }
}

Troubleshooting

Telemetry Not Appearing

1. Verify connection string is correct and accessible
2. Check network connectivity to Azure ingestion endpoint
3. Ensure TelemetryClient.Flush() is called before app shutdown
4. Check sampling configuration - you might be filtering too aggressively
5. Verify no firewall blocking outbound connections to Azure
6. Wait 2-3 minutes for telemetry to appear in portal

High Costs

1. Enable adaptive sampling to reduce volume
2. Filter unnecessary telemetry at the source
3. Review and adjust data retention policies
4. Use summary metrics instead of detailed events where possible

Missing Dependencies

1. Verify dependency tracking is enabled in configuration
2. Check that packages support automatic instrumentation
3. Manually track dependencies if auto-collection doesn't work

Environment Configuration

Development

{
  "ApplicationInsights": {
    "ConnectionString": "InstrumentationKey=dev-key",
    "EnableAdaptiveSampling": false,
    "EnableDebugLogger": true
  }
}

Production

{
  "ApplicationInsights": {
    "ConnectionString": "InstrumentationKey=prod-key",
    "EnableAdaptiveSampling": true,
    "SamplingPercentage": 5.0
  }
}

References

• Application Insights Overview
• ASP.NET Core Integration
• Node.js Integration
• Sampling in Application Insights
• Custom Events and Metrics