An AWS Region is the physical location where multiple clusters of data centers are deployed to provide different cloud services. A logical group of data centers is called the availability zone, and there is a minimum of 3 availability zones in a region. These availability zones are physically isolated from each other inside the region. Each availability zone inside the region has its own power, cooling, and physical security system and has redundant, ultra-low latency networking. In order to run highly available applications on AWS, design the application to run in multiple availability zones. In case one availability zone goes down, your application will still be running on another availability zone.
AWS has the most extensive global infrastructure as compared to other popular cloud providers and serves cloud services to customers worldwide. Currently, AWS is maintaining almost 21 regions in the following geographical areas
- Middle East
- South Africa
- Asia Pacific
- South America
- North America
An availability zone collects one or more data centers with ultra-low latency and redundant networking inside a region. There is a minimum of three availability zones inside a single AWS region. An application can be deployed across multiple availability zones in a region for high availability. In case one availability zone goes down, the application will still be accessible in the second availability zone.
All the availability zones inside a region are connected to each other with high bandwidth, ultra-low latency network that makes the communication between the availability zones faster. Each availability zone has a redundant power supply, networking, and connectivity and the traffic between these availability zones is encrypted. The network performance is fast enough for synchronous replication across availability zones.
If an application is deployed across multiple availability zones, your application will be safe from natural disasters like floods, earthquakes, etc. The average distance between two availability zones is considerable, and all the availability zones lie within the 100km radius.
AWS local zones are used to place different AWS services like compute, storage, and databases closer to the end-users. AWS local zones, real-time gaming, media, and entertainment applications requiring single-digit milliseconds latency can be run closer to the end-users.
AWS local zones can be used to run the low latency applications using AWS services like Amazon EC2, Amazon VPC, Amazon EBS, Amazon EFS, and Amazon ELB. Local zones allow the end-users to connect to the services seamlessly by providing a high bandwidth and secure connection between AWS and local zones workloads. The main use cases of local zones are listed below.
- Running low latency applications at the edge
- Meeting the stringent data residency requirements
- Simplifying hybrid cloud migrations
AWS edge locations
Edge locations in AWS are the data centers located near the end-users and designed to deliver content from different AWS services with minimum latency. Amazon has a lot of edge locations located around the globe. When a user makes a query, AWS first checks if the data is available on the edge location. If the data is available on the edge location, it is delivered to the end-user without making a query to the AWS servers. Otherwise, data is fetched from the AWS servers and delivered and cached at the edge location.
Following are some services that use AWS edge locations for faster content delivery.
- CloudFront: CloudFront caches the copy of data at edge locations, so the content is saved closer to the user. This content then can be delivered to the end-users faster.
- Web application firewall: A web application firewall is used at the edge locations to filter unwanted data earlier before it reaches the AWS region.
- Route53: Route53 is a DNS server and is used to serve the DNS queries from the edge locations. When a DNS query is made, it is cached at the edge location, and the next time, it returns the response from the edge location.
AWS direct connect locations
In order to connect your local data center to the AWS so that your local services and AWS services can talk to each other privately, AWS provides direct connect locations. AWS direct connect locations to connect your private network to AWS direct connect locations over fiber optic cables. With AWS direct connect, a virtual interface is created to the AWS public services like Amazon VPC, Amazon S3, directly bypassing the ISPs in the network path. AWS direct connect provides access to the public services in the region in which the direct connect is deployed.
AWS global infrastructure contains more than 100 direct connect locations around the globe to provide a better hybrid environment. Following is the list of geographical regions in which the AWS direct connect is located.
- Asia Pacific
- Middle East
- South America
- United States
Each of the above geographical regions contains multiple AWS direct connect locations.
AWS provides over 200 services in different geographical areas across the globe. AWS manages regions, availability zones, direct connect locations, and local zones to provide better services to the end-users. Before deploying any application, it is good to have a basic understanding of the global infrastructure of AWS. This blog describes detailed information and structure of the AWS global infrastructure.