Azure Application Insights

Install Packages

dotnet add package Microsoft.ApplicationInsights.AspNetCore

Configure

Program.cs

using Microsoft.ApplicationInsights;

var builder = WebApplication.CreateBuilder(args);

// Enable Application Insights
builder.Services.AddApplicationInsightsTelemetry(
    builder.Configuration["ApplicationInsights:ConnectionString"]);

builder.Services.AddSingleton<WeatherService>();

var app = builder.Build();

app.MapGet("/", () => "Hello World with App Insights!");

app.MapGet("/weather", (WeatherService service) =>
{
    service.GetWeather();
    return Results.Ok("Weather checked and telemetry sent!");
});

app.Run();

public class WeatherService
{
    private readonly TelemetryClient _telemetry;

    public WeatherService(TelemetryClient telemetry)
    {
        _telemetry = telemetry;
    }

    public void GetWeather()
    {
        // Track custom event
        _telemetry.TrackEvent("GetWeatherCalled");

        // Track custom metric
        _telemetry.GetMetric("WeatherRequests").TrackValue(1);

        // Simulate error
        try
        {
            throw new Exception("Test Exception from WeatherService!");
        }
        catch (Exception ex)
        {
            _telemetry.TrackException(ex);
        }
    }
}

Add ConnectionString

appsettings.json

{
  "ApplicationInsights": {
    "ConnectionString": "InstrumentationKey=xxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx;IngestionEndpoint=https://<region>.in.applicationinsights.azure.com/"
  },
  "Logging": {
    "LogLevel": {
      "Default": "Information",
      "Microsoft.AspNetCore": "Warning"
    }
  },
  "AllowedHosts": "*"
}

• Visit /weather → This will trigger events, metrics, and a sample exception.

• Open Azure Portal → Application Insights → Logs / Live Metrics.

• You’ll see your requests, events, and errors streaming in real-time. 🎉

Querying Telemetry with KQL

-Application Insights uses Kusto Query Language (KQL) for analyzing telemetry data.
For example,

see all failed requests:

requests
| where success == false
| order by timestamp desc

find the top 5 slowest endpoints

requests
| summarize avg(duration) by name
| top 5 by avg_duration desc