AWS Cloud — what is it physically?
It is a connected network of AWS regions. This map shows the underwater cabling between the AWS regions.
This is what an AWS region is like physically.
A campus of buildings,
where each building has multiple stacks of servers,
And each server has hypervisor running on it so that the server comes a group of VMs (EC2s)
EC2 is the building block of AWS infrastructure/ AWS Cloud.
This is already in place — a grand scheme. And then you come along with your business plan, and provision an EC2 here and an EC2 there, and run your applications.
AWS has 24 plus regions and 77 plus availability zones.
Region is a location, and it can have multiple availability zones.
Availability zone is a set of data centers that are very close to each other.
Data center is a building with rows and rows of servers.
Edge locations are locations used by AWS for caching content.
There are more edge locations than there are regions around the world.
Shared responsibility model
When you use resources from AWS, there are certain things that the customer is responsible for, and certain things that AWS is responsible for. Typically, the underlying infrastructure would be AWS responsibility, and the services on top of the infrastructure would be the customer’s responsibility.