AWS Fundamentals for Networking

AWS infrastructure has a hierarchy in regards to the size, scope and function of each component.

Regions

Largest part of the infrastructure. They are based on geographical location. They are completely isolated from other regions and should be completely separate failure domains. They consist of multiple availability zones. New Regions are being built all the time.

Region names are based on area they are in ‘us-west-1’, ‘us-east-2’, etc. These are key names that you will reference when making API calls and accessing the AWS console.

AWS GOV cloud is provided for governmental agencies and other enterprise that might have more highly confidential data or critical functions.

Availability Zones

These are data centers or at the least separate failure domains within a region that offer AWS services. An AZ could be made up of multiple data centers in an area. They are supposed to be isolated from other availability zones within the same region. Usually there are 2-5 AZ’s per region.

AWS states that the latency between these AZ’s should be 2ms or less. I Need to research if this is one way or RT?

Each AZ has a letter append to the region name to identify it: ‘us-west-1a’, ‘us-east-2c’, etc. This naming system is based on account. So what could be the ‘a’ AZ in one account for a region could be completely separate hardware and a different physical AZ then what another account has listed for ‘a’. This is done for load balancing purposes.

AZ names are mapped to an AZ ID.

Edge Locations

Make up CDN networks. Provides cached information services. DNS(Route 53) and AWS CloudFront are two examples. Low latency data as close to the request source as possible.

Infrastructure.aws.com is good resource to see all of the Edge Locations, Regions and Availability Zones.

AWS Workloads and Services.

Within an AZ you almost certainly use a VPC(Virtual Private Cloud) The services you will run here will typically be your own workloads such as EC2 instances.

There are also many AWS public services that run outside of your own infrastructure on dedicated resources provided by AWS:

• S3

• CloudWatch

• CloudFormation

• DynamoDB

These will typically be access through the internet gateway object.

Most resources will generally fall under three different functions: