Ref: https://learn.cantrill.io/courses/1820301/lectures/41301437

YouTube: https://www.youtube.com/watch?v=WFiMqGDUefw

• Summary:
• Contents:

Global and Regional Applications

• 🔧 Global applications (e.g. Netflix) consist of multiple regional applications
  • Regional applications = discreet blocks of infrastructure that operate independently & duplicated across different regions
  • Efficient concept → makes designing large platforms much easier
• 3 common types of system architectures in AWS:
  1. Small scale or single-region systems → only exist in 1 region
  2. Single-region systems with a DR requirement → failover/secondary region is needed
  3. Multi-region systems → need to operate through failure in 1 or more regions
• AWS products and services map onto specific architectural components
  • 💡 Note that most AWS products & services are regional, very few are actually global
    • All the regions make up the AWS global infrastructure

Architectural Components of Cloud Applications and Systems

Global Components

1. Global Service Location & Discovery
   • How is the app discovered? What happens when you type netflix.com in your browser?
   • e.g. DNS configuration in Amazon R53
2. Content Delivery (CDN) & Optimization
   • How does the content/data for an app get to users globally?
   • Are there pockets of storage distributed globally? Or does it all come from a central location?
   • e.g. Amazon CloudFront (CF)
     • Caches content for users in different locations
3. Global Health Checks & Failover
   • Monitor health of infrastructure in different locations
     • If infrastructure is unhealthy in a location, move customers to another location
   • Isolate failures and move sessions to a different region when they happen (failover)
   • e.g. R53 health checks

• Simplified example of Netflix global architecture (Primary region = US, Secondary = AUS)

Regional Components

1. Regional Entry Point
   • How do customers enter regional infrastructure?
   • e.g. through a VPC, or using public space AWS services
2. Scaling & Resilience
   • How does the system tolerate regional load changes and failures?
   • e.g. Application Load Balancers (ALBs), Auto Scaling Groups (ASGs)…
3. Application Services & Components
   • Functionality of the system in the region
   • e.g. EC2 instances, S3 buckets…