Global Accelerator

Anycast IP addresses and edge locations

AWS Global Accelerator utilizes anycast IP addresses, a special IP address that can be used by multiple devices simultaneously. Unlike traditional unicast IP addresses, anycast IP addresses can simultaneously be assigned to multiple devices or network interfaces. This unique feature allows multiple devices to share the same IP address across different locations. Each Global Accelerator edge location has this pair of anycast IP addresses. When traffic is destined for any of these anycast IP addresses, it can be serviced by any Global Accelerator edge locations.

When a user sends a request to an anycast IP address, the internet routing infrastructure automatically directs the request to the nearest available anycast node or edge location. This routing decision is typically based on factors such as network topology, latency, and routing protocols.

By leveraging anycast IP addresses, AWS Global Accelerator ensures that users are automatically routed to the nearest edge location, optimizing data delivery and reducing latency for improved performance.

Optimizing network connectivity

AWS Global Accelerator effectively moves the AWS network closer to the customers by employing anycast IP addresses and strategically distributing them across edge locations. This approach minimizes the distance data needs to travel, reducing latency and improving performance. Although this part of the connection still encounters variability inherent in the public internet, it is limited to the segment between the customer and the Global Accelerator edge location.

Transit over AWS global network

Once the data reaches any Global Accelerator edge locations, it is transited over the AWS global network to reach its destination. AWS’s dedicated global network offers reliable, high-performance connectivity, unlike the public internet, which is subject to external factors and varying performance. With fiber links connecting all AWS regions and complete control over network capacity and performance, AWS ensures optimal data transit between edge locations and destinations.

Comparison with CloudFront

The key difference between CloudFront and Global Accelerator is that Global Accelerator is a network product. It works on TCP or UDP applications, including web apps, whereas CloudFront caches HTTP and HTTPS content. Global Accelerator doesn’t cache anything and it doesn’t provide any HTTP or HTTPS capabilities, and it doesn’t understand the protocol for either of these. So we can conclude that:

• Both CloudFront and Global Accelerator aim to bring AWS’s network closer to users but serve different purposes.

• CloudFront caches HTTP/HTTPS content for faster delivery, while Global Accelerator optimizes network routes for TCP/UDP applications.

Benefits of AWS Global Accelerator

• Enhanced reliability and performance: By leveraging anycast IP addresses and the AWS global network, Global Accelerator reduces latency and ensures efficient data transit.

• Improved user experience: Users are automatically routed to the nearest edge location, minimizing the distance data needs to travel.

• Network control: AWS maintains complete control over its global network, ensuring consistent performance and capacity management across regions.