MSDN Documentation

Azure App Service

Monitoring Azure App Service

Effective monitoring is crucial for understanding the health, performance, and usage of your Azure App Service. This guide covers the key monitoring tools and techniques available within the Azure portal and through integrated Azure services.

Azure Portal Monitoring Tools

The Azure portal provides a centralized dashboard for monitoring your App Service instances. Key features include:

  • Metrics: View real-time and historical performance data such as CPU percentage, memory usage, requests, and data in/out.
  • Logs: Access application logs, web server logs, and diagnostic logs to troubleshoot issues.
  • Alerts: Configure alerts based on specific metric thresholds or log events to be notified of potential problems.
  • Application Insights: Integrate with Application Insights for deep application performance monitoring (APM), including request tracing, dependency mapping, and exception analysis.

Key Metrics to Monitor

Pay close attention to the following metrics:

  • HTTP Server Errors: HTTP 5xx and HTTP 4xx errors can indicate application or configuration issues.
  • CPU Percentage: Consistently high CPU usage may signal performance bottlenecks or resource exhaustion.
  • Memory Working Set: High memory consumption can lead to slow performance or application crashes.
  • Requests: Monitor the volume of incoming requests to understand traffic patterns and potential load issues.
  • Data In/Out: Track network traffic to identify unexpected data transfer or potential cost implications.

Application Insights Integration

Application Insights is Azure's extensible APM service that provides powerful capabilities for monitoring live applications. By integrating Application Insights with your App Service, you can:

  • Track Performance: Monitor request rates, response times, and failure rates.
  • Detect Anomalies: Automatically detect unusual behavior and potential issues.
  • Diagnose Failures: Analyze exceptions and trace requests to pinpoint root causes of errors.
  • Understand User Behavior: Gain insights into how users interact with your application.

To enable Application Insights, navigate to your App Service in the Azure portal, go to the 'Application Insights' blade, and follow the setup instructions.

Pro Tip:

Configure autoscale rules based on key metrics like CPU percentage to automatically adjust the number of instances running your App Service, ensuring performance and cost-efficiency.

Diagnostic Logs

Diagnostic logs provide detailed information about your App Service. You can configure the types of logs to collect, including:

  • Application Logs: Logs generated by your application code.
  • Web Server Logs: Logs from the underlying web server (e.g., IIS).
  • Detailed Error Messages: HTTP status code details for failed requests.
  • Failed Request Tracing: Detailed traces for specific failed requests.

These logs can be sent to various destinations, such as Azure Storage, Event Hubs, or Log Analytics, for further analysis and retention.

Note:

Enable Advanced Tools (Kudu) for deep diagnostics, including access to file system, process explorer, and log streaming.

Alerting

Setting up alerts is a proactive approach to monitoring. You can create alert rules in the Azure portal based on specific metrics or log search queries. When an alert condition is met, you can trigger actions like sending an email, SMS, or calling a webhook.

Example Alert Rule:

  • Condition: CPU Percentage > 80% for 15 minutes
  • Action: Send email to operations@example.com

Further Reading