Skip to main content

🔎 Deep Dive: Azure Application Insights — Monitoring & Integration with Azure Services

When you deploy apps to the cloud, monitoring and observability are critical. Azure Application Insights (App Insights) is a powerful tool that helps you monitor, diagnose, and optimize your applications.

In this blog, we’ll cover:

✅ What Azure App Insights is
✅ Key concepts (Telemetry, KQL, Instrumentation Key vs Connection String, Sampling)
✅ Integration with various Azure services
✅ Real-world setup examples (App Service, Function App, API, Containers, Key Vault, Cosmos DB, Service Bus)
✅ Best practices for developers

🌟 What is Azure Application Insights?

Azure Application Insights is an Application Performance Monitoring (APM) service under Azure Monitor.

It allows you to:

  • Collect telemetry: requests, dependencies, exceptions, traces, custom events
  • Diagnose performance issues (slow requests, failed calls, memory leaks)
  • Visualize health in dashboards
  • Write powerful queries using Kusto Query Language (KQL)
  • Integrate with DevOps pipelines (alerts, dashboards, release health)

Think of it as “X-ray vision for your app” 🚀.

🏗️ Key Concepts

1. Telemetry

The data collected:

  • Request telemetry → incoming HTTP calls
  • Dependency telemetry → calls to SQL, Cosmos DB, REST APIs
  • Exception telemetry → handled/unhandled errors
  • Trace telemetry → logging events (info/warning/error)
  • Metric telemetry → performance counters (CPU, memory)

2. Instrumentation Key / Connection String

  • Instrumentation Key → Old way (GUID) to connect your app to App Insights.
  • Connection String → New recommended way; includes Instrumentation Key + Endpoint info.

3. Sampling

  • Reduces telemetry volume (e.g., log 10% of requests).
  • Adaptive Sampling automatically adjusts based on traffic.

4. KQL (Kusto Query Language)

  • Query language to explore telemetry.
  • Example:
requests
| where success == false
| summarize count() by resultCode

⚙️ Integration with Azure Services

App Insights integrates deeply with multiple Azure services. Let’s explore one by one.

🔹 1. Azure App Service (Web Apps, APIs)

  • Integration: Enable App Insights directly from App Service → Monitoring → Application Insights.
  • Automatically collects: requests, dependencies, exceptions.
  • Example: Enable via code (ASP.NET Core)
builder.Services.AddApplicationInsightsTelemetry(builder.Configuration["APPLICATIONINSIGHTS_CONNECTION_STRING"]);
  • Benefit: Full visibility into your web app’s performance.

🔹 2. Azure Function Apps

  • Native integration: Go to Function App → Application Insights and enable it.
  • Collects: function executions, failures, dependency calls.
  • Example: Add manually in host.json
{
  "logging": {
    "applicationInsights": {
      "samplingSettings": {
        "isEnabled": true,
        "maxTelemetryItemsPerSecond" : 5
      }
    }
  }
}
  • Benefit: Monitor serverless workloads (cold starts, retries, errors).

🔹 3. Azure Cosmos DB

  • Integration: Cosmos DB automatically emits diagnostic logs → forward to App Insights via Diagnostic Settings.
  • You can track query RU consumption, latency, throttling.
  • Example:
cosmosClientOptions.Diagnostics.IsTelemetryEnabled = true;
  • Benefit: Detect slow queries, RU bottlenecks, and failed calls.

🔹 4. Azure Key Vault

  • Key Vault access logs (e.g., secrets retrieval) can be sent to Log Analytics / App Insights using Diagnostic Settings.
  • Example telemetry:
    • Who accessed a secret
    • Failed authentication attempts
  • Benefit: Detect unauthorized access patterns.

🔹 5. Azure Service Bus

  • Integration: Add App Insights SDK in your .NET worker/processor that consumes messages.
  • Example:
services.AddApplicationInsightsTelemetryWorkerService();
  • Track: Queue/Topic message latency, dead-letter counts, failed deliveries.
  • Benefit: Monitor message-driven systems in real time.

🔹 6. Azure SQL Database

  • Track DB calls as dependencies automatically when using Entity Framework / ADO.NET.
  • Example:
services.AddApplicationInsightsTelemetry();
  • Benefit: See slow queries directly in App Insights performance blade.

🔹 7. Azure AKS / Containers

  • Use Azure Monitor for containers to forward telemetry into App Insights.
  • Collects: pod restarts, node CPU/memory, container logs.
  • Benefit: End-to-end view of microservices.

🔹 8. Azure API Management (APIM)

  • APIM diagnostic logs can be forwarded to App Insights.
  • Example: Enable in APIM → Diagnostics → Application Insights.
  • Benefit: Monitor API usage, latency, backend errors.

🔹 9. Azure Application Gateway + App Insights

  • Integration: Enable diagnostics in App Gateway → Send to App Insights.
  • Collects: WAF logs, request/response times.
  • Benefit: Detect malicious traffic patterns, latency issues.

🖥️ Example: ASP.NET Core API + App Insights Integration

Step 1: Add NuGet

dotnet add package Microsoft.ApplicationInsights.AspNetCore

Step 2: Configure in Program.cs

var builder = WebApplication.CreateBuilder(args);

builder.Services.AddApplicationInsightsTelemetry(
    builder.Configuration["APPLICATIONINSIGHTS_CONNECTION_STRING"]);

var app = builder.Build();

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

app.Run();

Step 3: View telemetry in Portal

  • Go to Application Insights resource → Logs.
  • Run KQL queries like:
requests
| where name == "GET /hello"
| project timestamp, duration, success

📊 Example: Function App with App Insights

host.json

{
  "logging": {
    "applicationInsights": {
      "samplingSettings": {
        "isEnabled": true,
        "maxTelemetryItemsPerSecond": 5
      }
    }
  }
}

Function Code

[FunctionName("HelloFunction")]
public static IActionResult Run(
    [HttpTrigger(AuthorizationLevel.Function, "get")] HttpRequest req,
    ILogger log)
{
    log.LogInformation("HelloFunction processed a request");
    return new OkObjectResult("Hello from Azure Function with App Insights!");
}

📌 Best Practices

  • ✅ Always use Connection String (not just Instrumentation Key).
  • ✅ Use sampling to avoid cost explosion in high-traffic apps.
  • ✅ Define custom telemetry for business events:
var telemetry = new TelemetryClient();
telemetry.TrackEvent("OrderPlaced", new Dictionary<string, string> { { "OrderId", "123" } });
  • ✅ Integrate App Insights with Azure Monitor Alerts for proactive monitoring.
  • ✅ Use Distributed Tracing to track requests across microservices.
  • ✅ Secure telemetry data with RBAC (role-based access control).

🎯 Conclusion

Azure Application Insights is not just a logging tool — it’s a full observability platform that integrates seamlessly with almost every Azure service: App Service, Function Apps, Cosmos DB, Service Bus, Key Vault, SQL DB, Containers, and more.

By setting up App Insights correctly, you gain:

  • End-to-end visibility
  • Faster root cause analysis
  • Proactive monitoring and alerts
  • Insights into both technical health and business events

🚀 If you’re running workloads in Azure, App Insights should be a default part of your architecture.


Comments

Post a Comment

Popular posts from this blog

🏗️ Deep Dive: Understanding Every Concept in Microsoft Entra API Onboarding for .NET Developers

When working with Microsoft Entra (formerly Azure Active Directory), you’ll hear terms like App Registration, Tenant, Client ID, Audience, Scopes, Roles, Tokens, OBO flow , and more. If you’re new, it can feel overwhelming. This guide breaks down every key term and concept , with definitions, examples, and how they connect when you onboard and consume a new API. 🔹 1. Tenant Definition : A tenant in Entra ID is your organization’s dedicated, isolated instance of Microsoft Entra. Think of it like : Your company’s identity directory. Example : contoso.onmicrosoft.com is a tenant for Contoso Ltd. 🔹 2. App Registration Definition : The process of registering an application in Entra to give it an identity and permission to use Microsoft identity platform. Why needed : Without registration, Entra doesn’t know about your app. What it creates : Application (Client) ID – unique identifier for your app Directory (Tenant) ID – your organization’s ID Types of apps : Web ...
Post a Comment
Read more

🗑️ Garbage Collection & Resource Management in .NET (C#) — Beginner Friendly Guide

When you start working with .NET and C#, one of the biggest advantages is that you don’t need to manually manage memory like in C or C++. The Garbage Collector (GC) does most of the work for you. But here’s the catch — not everything is managed automatically. Some resources like files, database connections, sockets, and native memory still need special handling. This blog will help you understand: ✔ How the GC works ✔ What are managed vs unmanaged resources ✔ The difference between Dispose , Finalize , and using ✔ The Dispose pattern with examples ✔ Best practices every C# developer should know 1) How Garbage Collection Works in .NET Managed resources → Normal .NET objects (string, List, etc.). GC frees them automatically. Unmanaged resources → External resources like file handles, database connections, sockets, native memory. GC cannot clean them up — you must do it. 👉 GC uses a Generational Model for performance: Gen 0 : Short-lived objects (local variables, t...
Post a Comment
Read more

☁️ Azure Key vault Short Notes

🟢 What is Azure Key Vault? A cloud service for securely storing and accessing secrets, keys, and certificates . Removes the need to keep secrets (like connection strings, passwords, API keys) inside code or config files. Provides centralized secret management, encryption, and access control . 👉 Think of it like a secure password manager but for your applications. 🟢 Key Features Secrets → store text values (e.g., DB connection string, API key). Keys → store cryptographic keys (RSA, EC) for encryption, signing. Certificates → store/manage SSL/TLS certificates. Access Control → Access Policies (older model). Azure RBAC (modern, preferred). Integration → works with App Service, Functions, AKS, VMs, SQL DB, etc. Logging → audit who accessed secrets via Azure Monitor / Diagnostic Logs. 🟢 Why Use Key Vault? Security → secrets are encrypted with HSM (Hardware Security Modules). Compliance → meet industry standards (PCI-DSS, ISO, GDPR). Automation → aut...
Post a Comment
Read more