Skip to content
What is AWS Global Accelerator?
AWS Global Accelerator is a networking service that improves the availability and performance of applications with global users by routing user traffic through Amazon Web Service’s vast global infrastructure.
It uses the AWS global network to direct traffic to the nearest and healthiest endpoint based on health checks and geographic location. This helps reduce internet latency, improve transfer speeds, and increase the reliability of application access.
Key points:
- AWS Global Accelerator supports two types of accelerators:
- Standard Accelerator: Improves the availability of your internet applications globally by directing traffic over the AWS network to the closest regional endpoint.
- Custom Routing Accelerator: Custom Routing Accelerators allow specific user traffic routing to be designated among multiple endpoints within a VPC subnet, targeting private IP addresses.
- Custom Global Accelerator routes traffic to the designated instance and port (image source: AWS)
- AWS Global Accelerator provides static IP addresses as fixed entry points for your applications, reducing the complexity associated with managing various IP addresses across different AWS Regions and Availability Zones. This setup avoids the need for client-facing changes or DNS record updates when modifying or replacing endpoints.
- It supports various endpoint types such as Network Load Balancers, Application Load Balancers, EC2 instances, and Elastic IP addresses across one or multiple AWS Regions.
- You can use it to dial traffic up or down for a specific AWS Region by configuring a traffic dial percentage for your endpoint groups. This is especially useful for testing performance and releasing updates.
- You can use it to control the proportion of traffic directed to each endpoint within an endpoint group by assigning weights across the endpoints.
- By default, AWS Global Accelerator offers two static IP addresses, or four addresses (two IPv4 and two IPv6) for dual-stack configurations. Users can also bring their own IP addresses (BYOIP). These IPs serve as consistent entry points for your applications and remain with your accelerator as long as it exists, even if the accelerator is disabled.
How AWS Global Accelerator works
#1: User initiates a request: When a user initiates a request, it’s directed towards AWS Global Accelerator’s static IP addresses, which serve as a fixed entry point to the applications.
#2: Traffic Routing: AWS Global Accelerator processes the request, leveraging the AWS global network. It assesses the health and geographic location of the nearest application endpoints to determine the most efficient way to route traffic.
#3: Endpoint Optimization: The service selects the optimal endpoint among configured options such as Network Load Balancers, Application Load Balancers, or EC2 instances across multiple AWS Regions. This decision is based on factors like endpoint health and geographical proximity to minimize latency.
#4: Request Processing and Response: The chosen endpoint processes the user request. The response is then routed back through the Global Accelerator to the user, optimizing the delivery path for speed/reliability.
Use cases for AWS Global Accelerator
Some common use cases for AWS Global Accelerator include:
Simpler network scaling for applications: As application usage increases, so does the demand for IP addresses and endpoints. AWS Global Accelerator helps you scale your network efficiently by associating your regional resources—like load balancers and EC2 instances—to static IPv4 and IPv6 addresses. Once listed in your client applications, firewalls, and DNS records, these IP addresses don’t need updating when you add or remove endpoints, conduct blue/green deployments, or run A/B tests. This feature is particularly valuable in sectors like IoT, retail, and healthcare, where frequent client updates are impractical.
Low-latency applications like VoIP, Gaming, Streaming, etc: For applications where low latency is critical—such as gaming and financial services—Global Accelerator increases performance by directing traffic to the nearest application endpoint, thus reducing internet latency.
Availability, disaster recovery and resiliency: Global Accelerator ensures your network remains available across multiple AWS Regions and independent network zones, boosting disaster recovery and compliance. It automatically reroutes traffic to a secondary region if the primary one fails. (If one static IP address becomes unavailable due to IP address blocking or unreachable networks, AWS Global Accelerator provides fault tolerance to client applications by rerouting to a healthy static IP address from the other isolated network zone.)
Security and DDOS prevention: Global Accelerator secures your applications by hiding AWS origins behind two static entry points, minimizing exposure to public internet threats. These entry points are shielded from DDoS attacks by AWS Shield and connect to your internal resources through a private, secure peering connection.
How to get started with AWS Global Accelerator
You can set up AWS Global Accelerator by using the API, the AWS Management Console, or an AWS CloudFormation template. Here are the three basic steps:
- Create an Accelerator: When you create your accelerator, AWS Global Accelerator allocates two static IP addresses that are immediately operational. You then set up one or more listeners to handle inbound connections from end clients, based on specified protocols and ports.
- Configure Endpoint Groups: Select one or more regional endpoint groups to connect to your accelerator’s listener, designating the AWS Regions where you wish to direct traffic. Your listener then forwards requests to the registered endpoints in these groups. AWS Global Accelerator continuously assesses the health of these endpoints through predefined health check settings. Additionally, you can manage the volume of traffic each endpoint group receives by adjusting the traffic dial percentage.
- Register Endpoints: In each endpoint group, register regional resources such as Application Load Balancers, Network Load Balancers, EC2 Instances, or Elastic IP addresses. You also have the option to assign weights to each endpoint, determining the distribution of traffic among them.
You can learn more about the practical steps in the AWS documentation.
Frequently Asked Questions
AWS Global Accelerator vs CloudFront
AWS Global Accelerator and Amazon CloudFront serve distinct, yet complementary, roles in managing application traffic. While Global Accelerator optimizes traffic routing to the nearest AWS endpoint to minimize latency and improve performance, CloudFront is a content delivery network (CDN) that caches content closer to users to reduce response times and bandwidth costs. Choosing between them depends on whether your priority is reducing latency for dynamic content with Global Accelerator or improving content delivery speeds with CloudFront.
AWS Global Accelerator vs Route53
How does AWS Global Accelerator pricing work
AWS Global Accelerator pricing is based on two main components: the number of accelerators you deploy and the volume of data transferred over those accelerators.
You pay a fixed hourly rate for each accelerator, which includes the benefit of having AWS-provided static IP addresses. Additionally, data transfer fees are calculated based on the amount of data processed by the accelerators, with prices varying by geographic region.
Understand and optimize AWS costs with nOps
If you’re looking to understand your AWS usage and costs, nOps can help.
nOps Business Contexts transforms millions of rows of contextless data into the who, what, when and why of cloud spend — making it easy to get 100% visibility of your cloud costs and usage so your bills are never a surprise or mystery.
Allocate 100% of your AWS costs, including EKS. Kubernetes costs are often a black box — no longer with nOps. Understand and allocate your unified AWS spend in one platform.
Automated resource tagging. You don’t need to have all your resources tagged to allocate costs. Create dynamic rules by region, tags, operation, accounts, and usage types to allocate costs back to custom cost centers.
40+ views & filters. Map hourly costs by any relevant engineering concept (deployment, service, namespace, label, pod, container…) or finance concept (cost unit, purchase type, line item, cost allocation tag…).
Custom reports & dashboards for the whole team. Monthly reporting and reconciliation can take hours; with nOps only minutes. Tailor dashboards and Slack/email reports to your needs, whether you’re a CFO or VP of Engineering.
The best part? nOps is an all-in-one solution for all of your cloud optimization needs: automated commitment management, rightsizing, resource scheduling, workload management, Spot usage, storage optimization, and more. Join our customers using nOps to understand your cloud costs and leverage automation with complete confidence by booking a demo today!
