Visual overview
Global Accelerator brings traffic onto the AWS global network and steers it toward healthy Regional application endpoints.
Technical reference
Global Accelerator combines static anycast entry points with health-aware Regional endpoint selection.
AWS currently assigns two static IPv4 addresses to a standard IPv4 accelerator.
Listeners define accepted protocol and port ranges.
Groups supported endpoints and applies a traffic dial at the Regional group level.
Health affects eligibility; weights influence traffic among endpoints in a group.
Service limits and capabilities can change. Values shown here reflect the current AWS documentation; use the linked official sources below as the source of truth.
A global static entry point onto the AWS network
AWS Global Accelerator improves availability and performance for supported internet-facing applications by providing static anycast IP addresses announced from the AWS edge network. Client traffic enters the AWS global network close to the user and is routed toward healthy application endpoints according to accelerator configuration, endpoint health and network conditions. This reduces reliance on long variable internet paths between the user and the destination Region.
For a standard IPv4 accelerator, AWS currently allocates two static IPv4 addresses. Dual-stack configurations include IPv4 and IPv6 static addresses according to the current service behavior. Treat the exact address count as a current service value rather than a universal networking principle; the durable exam concept is static anycast entry points backed by the AWS global network.
Listeners, endpoint groups and traffic controls
A standard accelerator has listeners for TCP, UDP or both on configured ports. Endpoint groups are associated with AWS Regions and contain supported application endpoints such as Application Load Balancers, Network Load Balancers, EC2 instances or Elastic IP addresses. Health checks and endpoint health determine which destinations are eligible to receive traffic.
Traffic dials control the percentage of accelerator traffic sent to an endpoint group, while endpoint weights influence the proportion delivered among endpoints inside the group. These controls enable operational patterns such as gradual Regional traffic shifts or temporarily draining a Region. They are routing controls, not application autoscaling controls: capacity behind the endpoint still has to meet the traffic it receives.
Global Accelerator versus CloudFront and Route 53
CloudFront is a CDN that can cache HTTP content at edge locations and reduce origin requests. Global Accelerator does not cache application content; it optimizes the network path for supported TCP/UDP traffic and supplies stable anycast IP entry points. This makes it useful for applications that need global performance and failover but are not primarily solved by caching, including certain APIs, gaming, voice or other latency-sensitive network traffic.
Route 53 can also route users across Regions using DNS policies, but DNS returns answers that clients and resolvers may cache for a TTL. Global Accelerator keeps the client-facing anycast IPs stable while AWS changes the selected healthy backend path. The services can be used together in a broader architecture, but exam scenarios often distinguish DNS-based routing, cached content delivery and anycast network acceleration.
Key takeaways
- 01
Global Accelerator provides static anycast IP entry points and routes supported traffic across the AWS global network.
- 02
Listeners define protocol/ports; endpoint groups organize Regional endpoints and health-based routing.
- 03
Traffic dials operate at endpoint-group level; weights distribute traffic among endpoints in a group.
- 04
CloudFront caches HTTP content; Global Accelerator accelerates supported TCP/UDP network paths without caching content.
Official AWS sources
Use these primary AWS resources for the source material behind this article and for deeper reference.