Pillars of a great Azure architecture

Security

• Defense in depth

  • Data

  • Application

  • VM/compute

  • Networking

  • Permiter

  • Policies & access

  • Physical Security

Performance and scalability

• Scaling

  • Scale Up (1 small VM -> 1 big VM)

  • Scale Out (1 small VM -> many small VMs)

• Patterns

  • Data partioning

  • Caching

  • Auto Scaling

  • Decouple resrouce-intensive tasks to as background jobs

  • Use a messaging layer between services

  • Implement scale units

  • Performance monitoring

Availability and recoverability

• RPO (Recovery Point Objective): The maximum duration of acceptable data loss

• RTO (Recovery Time Objective): The maximum duration of acceptable downtime

Efficiency and operations

• Efficiency Best practices:

  • Cost optimization

  • move from IaaS to PaaS where possible

• Operation Best practices:

  • Automate

  • Monitoring, Logging, Instrumentation

  • design with DevOps in mind: infrastructure as code, automated testing and deployment

  • Eliminate waste