Skip to content
Pablo Rodriguez

Global Infrastructure

  • AWS Global Infrastructure designed for:

    • Flexibility
    • Reliability
    • Scalability
    • Security
    • High-quality global network performance
  • For current infrastructure information:

    • AWS Global Infrastructure Map: aws.amazon.com/about-aws/global-infrastructure
    • Regions and Availability Zones: aws.amazon.com/about-aws/global-infrastructure/regions_az/
  • AWS has 22 Regions worldwide
  • Each Region:
    • Physical geographical location
    • Contains one or more Availability Zones
    • Regions are isolated from one another
      • Resources not automatically replicated across Regions
      • “It is your responsibility to replicate data across Regions”
    • Regions introduced after March 20, 2019 are disabled by default
    • Some Regions have restricted access:
      • Amazon AWS (China) - Beijing and Ningxia Regions only
      • AWS GovCloud (US) - For US government agencies
  • Factors to consider:
    • Data governance and legal requirements
      • Local laws might require information kept within geographical boundaries
    • Proximity to customers (latency)
      • Run applications/store data close to users
      • Can test latency using CloudPing (cloudping.info)
    • Services available within Region
      • Not all services available in all Regions
    • Costs (vary by Region)
      • Example: t3.medium EC2 instance
        • US East (Ohio): $0.0416/hour
        • Asia Pacific (Tokyo): $0.0544/hour
  • Each Region has multiple Availability Zones
  • Each Availability Zone:
    • Fully isolated partition of AWS infrastructure
    • Consists of discrete data centers
    • Designed for fault isolation
    • Has own power infrastructure
    • Physically separated (but within 100 km of other AZs)
    • Interconnected with high-speed private networking
      • High-bandwidth, low-latency
      • Synchronous replication between zones
    • You choose your Availability Zones
    • AWS recommends replicating across zones for resiliency
  • Foundation for AWS infrastructure
  • Customers don’t specify data centers (AZ is most granular specification)
  • Features:
    • Secure design
    • Redundant design to tolerate failure
    • Critical systems backed up across multiple AZs
    • Continuous monitoring for capacity planning
    • Locations not disclosed
    • Automated processes move traffic away from affected areas
    • Custom network equipment from multiple ODMs
    • Each typically has 50,000 to 80,000 physical servers
  • Amazon CloudFront: content delivery network (CDN)
  • Amazon Route 53: Domain Name System (DNS) service
  • Edge locations:
    • Requests routed to nearest edge location to lower latency
    • Located in most major cities around the world
  • Regional edge caches:
    • Used by default with CloudFront
    • For content not accessed frequently enough to remain in edge location
  • Elasticity and scalability
    • Dynamic adaptation of capacity
    • Adapts to accommodate growth
  • Fault-tolerance
    • Continues operating despite failed components
    • Built-in redundancy
  • High availability
    • Minimal downtime
    • Minimal to no human intervention
  • AWS Global Infrastructure consists of Regions and Availability Zones
  • Region choice typically based on compliance requirements or to reduce latency
  • Each Availability Zone physically separate with redundant power, networking, connectivity
  • Edge locations and Regional edge caches improve performance by caching content closer to users

The AWS Global Infrastructure provides the foundation for AWS services, delivering a reliable, scalable, and secure environment with global reach and local presence.