Scaling and troubleshooting Web Apps

Azure App Services

Florin Angelescu

Azure Cloud Architect

Why scaling matters

 

 

• Scaling up
  • More resources (CPU, memory, etc.)

• Scaling out
  • More instances

Manual vs. automatic scaling

 

 

• Manual scaling
  • Control the number of instances directly

• Auto-scale
  • Adjusts resources automatically
  • Based on rules you define

Configuring auto-scale rules

 

• Add an instance if CPU usage stays above 70% for 10 minutes
• Set limits to avoid provisioning too many instances

Monitoring your Web App

• Monitoring:
  • Helps detect issues early
  • Supports informed decisions
• Azure provides built-in metrics such as:
  • Response time
  • Request rates
  • Error counts
• Metrics can be viewed in real time
• Basic metrics offer good insights
• Deeper analysis is needed for effective troubleshooting

What is Application Insights?

Enabling Application Insights

 

 

• Enable Application Insights
  • Directly in Azure Portal
• Collects data automatically
  • No need for code changes

Troubleshooting example

• Use Application Insights to check the "Performance" tab
  • Response times increase sharply during those hours
• Look at "Dependencies"
  • A database call is taking too long
• Optimize the database query
  • Improves your application speed

Summary

 

 

• Scale out
  • Add more instances to handle increased load
• Scale up
  • Increase compute resources for faster performance
• Monitor
  • Track performance and troubleshoot issues

Let's practice!

Azure App Services