Networking and Content Delivery
AWS Global Accelerator
AWS Global Accelerator directs traffic through the AWS global network to healthy regional endpoints using static anycast IP addresses. It improves availability and network consistency for non-cacheable global GenAI APIs.
Key points
- Endpoint groups control regional traffic dials
- Health checks shift new connections away from unhealthy endpoints
- Anycast addresses remain stable as endpoints change
When to use it
- Route users to healthy regional inference gateways
- Provide fixed IP allowlisting for a multi-Region AI API
Exam tips
- Choose Global Accelerator for dynamic TCP or UDP application traffic and CloudFront for cacheable HTTP content and edge processing
- Traffic steering must still honor model availability and data-residency rules