Ref: https://learn.cantrill.io/courses/1820301/lectures/41301618 and https://www.udemy.com/course/aws-certified-cloud-practitioner-new/learn/lecture/20054392

• Summary
• Contents:

AWS Global Network

• AWS infrastructure is spread globally throughout the world, divided in collections of smaller groupings of infrastructure connected together with high speed global NWing
  • Ref: https://www.infrastructure.aws/ → view of the AWS global NW
  • 💡The AWS global NW is constantly growing, evolving and changing
• 3 types of AWS infrastructure groupings:
  1. Regions
  2. Availability Zones (AZs)
  3. Edge Locations = Points of Presence (PoPs)

AWS Infrastructure Groupings

AWS Region

• 🔧 Geographical area with full AWS infrastructure (compute, storage, DB, AI, analytics…)
  • E.g. The Asia Pacific (Sydney) region, or ap-southeast-2 region
    • “Asia Pacific (Sydney)” = region name ↔ ap-southeast-2 = region code
  • ‼️ Does not equate to states, countries nor continents!
    • The region area is selected/created by AWS to deploy their infrastructure
• Regions are interconnected with each other at high speeds
• Regions allow design of products that can withstand global disasters
  • Even if a whole region has an outage, products can run from a different region
• A region must be specified when interacting with a non-global AWS service
  • Global services (e.g. IAM) do not need to specify a region
• 🔧 3 main benefits:
  1. 👍 Geographic separation → Isolated Fault Domain (stability & resilience)
     • If a region has an outage, products will only experience outage in that region
  2. 👍 Geopolitical separation → Different Governance
     • Data in a region is affected by the laws of the country where the region is located
     • Data does NOT move between regions unless you configure/allow AWS to do so
  3. 👍 Location control → Performance
     • Deploy product closer to users → lower latency and higher performance
• Considerations when choosing region(s) for your app/infrastructure
  • Compliance to data regulations → e.g. if data generated in EU must stay in EU
  • Latency → infrastructure closer to end users means lower latency, better performance, better user experience (UX)
  • Available services → Not all services are supported in all regions!
    • Service availability per region: https://aws.amazon.com/about-aws/global-infrastructure/regional-product-services
  • Pricing → certain regions will have different pricing for certain services (but most services have a similar pricing no matter the region)

AWS Availability Zone (AZ)

• 🔧 AWS regions are divided into AZs
  • ❗ 3-6 AZs per region
  • e.g. The ap-southeast-2 region is divided into the ap-southeast-2a, ap-southeast-2b and ap-southeast-2c AZs
  • AZs provide isolation within a region
    • Compute, storage, NWing, power & facilities isolation
    • A service can fail in an AZ, but (if configured to do so) can continue to run in other AZs that are intact → AZ-resilient service