When beginning with AWS, the concepts of Regions, Availability Zones (AZs), Data Centers and Edge Locations can be a little confusing. Let’s review some facts that will help clarify the relationship.
First of all, they are listed in this title in their relative order of size.
For the moment, let’s remove Edge Locations from consideration. Their service offerings differ slightly from Regions, AZs and Data Centers. This is an important distinction.
An AWS Region is the largest unit.
- A Region is a geographic location in the world. As of this writing (July 2017), there are sixteen Regions across the globe with 44 AZs.
- Given the relationship above, it’s apparent that a Region consists of multiple Availability Zones (AZs). For example, São Paulo, a Region, has 3 Availability Zones. (For many people, GovCloud (US) and Beijing (China) are Regions but are not available, reducing their available Regions to fourteen.)
- You can see available regions from the CLI by typing the command below. Note that you may not see GovCloud or Beijing in your list!
aws ec2 describe-regions
- An Availability Zone consists of one or more discrete data centers.
- As we saw above with Regions, you can view a list of Availability Zones using the CLI.
aws ec2 describe-availability-zones --region eu-west-2
- A Data Center is the smallest unit of consideration among the three (Regions, Availability Zones and Data Centers.) An important distinction is that the discrete data centers may not necessarily be spread by great distance physically, but will have completely separate power, etc.
- According to the 2014 article, Amazon AWS Regions vs Availability Zones vs Edge locations vs Data centers, “a data center typically have 50,000 to 80,000 physical servers”.
There will be many more Edge Locations. For example, Chicago hosts Edge Locations, but not a Data Center. The primary purpose of an Edge Location is twofold:
- CDN or Content Delivery Network for cached content via CloudFront
- DNS via Route 53