One of the most important aspects of running your application in an efficient and highly available manner is the correct selection of the location of your AWS services. AWS is designed to offer multiple facilities in several different regions so that it can provide both high availability of services within a single region and the ability for us to architect solutions that are distributed and replicated across different regions. 

Based on the experience gained from running Amazon since the 90s, AWS has set out to create a world-class infrastructure architecture that is composed of the following components:

• Datacenter
• Availability zones
• Region

The lowest infrastructure component in the AWS architecture is the datacenter. A datacenter is essentially designed to hold raw compute, network, and storage capacity and to provide no redundancy to your applications at all. The AWS datacenter design has the following characteristics:

• Compute capacity: Between 50.000 – 80.000 servers
• Storage capacity: Approximately 11 petabytes per rack
• Network throughput: Up to 100 terabits per second
• Proprietary redundant network stack, delivered via multiple other device manufacturers
• Proprietary redundant network security stack: