Each listener is associated with an endpoint group, and each endpoint group contains one or more endpoints.

Endpoint groups

Each endpoint group is associated with a specific region. You can associate an endpoint group with a listener by specifying the region to which you want to distribute network traffic. After you associate an endpoint group with a listener, the system distributes network traffic to the optimal endpoints in the endpoint group.

Listeners that use different protocols support different types of endpoint groups:
  • TCP or UDP listeners

    By default, you can create two default endpoint groups for each TCP or UDP listener. If you want to create more default endpoint groups, go to the Quota Management page and increase the quota of gaplus_quota_epgs_per_listener. For more information, see Manage quotas.

    You must deploy default endpoint groups in different regions. You can set a traffic distribution ratio for each default endpoint group. The traffic distribution ratio specifies the proportion of traffic that is distributed to a default endpoint group.

  • HTTP or HTTPS listeners

    By default, you can create one default endpoint group and one virtual endpoint group for each HTTP or HTTPS listener. If you want to create multiple virtual endpoint groups, go to the Quota Management page and increase the quota of gaplus_quota_vepg_per_listener. For more information, see Manage quotas.

    • A default endpoint group refers to the endpoint group that you configure when you create an HTTP or HTTPS listener.
    • A virtual endpoint group refers to the endpoint group that you can create on the Endpoint Group page after you create a listener.

      After you create a virtual endpoint group for an HTTP or HTTPS listener, you can create a forwarding rule and associate the forwarding rule with the virtual endpoint group. Then, the HTTP or HTTPS listener forwards requests with different destination domain names or paths to the default or virtual endpoint group based on the forwarding rule. This way, you can use one Global Accelerator (GA) instance to accelerate multiple domain names or paths. For more information about how to create a forwarding rule, see Create and manage forwarding rules.

Note You can create endpoint groups for basic GA instances without the need to create listeners. A basic GA instance supports only one endpoint group and one endpoint.

Endpoints

Endpoints are destinations of client requests. You can add at most four endpoints to an endpoint group. The following table describes the backend service types of endpoints.
Backend service area Network type Backend service type Backend service
Alibaba Cloud Internet Alibaba Cloud public IP address
  • Elastic IP addresses (EIPs)
  • Static public IP addresses

    Static public IP addresses include the public IP addresses of Elastic Compute Service (ECS) instances and the public IP addresses of Internet-facing Classic Load Balancer (CLB) instances that are deployed in classic networks.

CLB Internet-facing CLB instances
VPC ECS ECS instances

Only ECS instances that are deployed in virtual private clouds (VPCs) are supported.

CLB Internal-facing CLB instances

Only CLB instances that are deployed in VPCs are supported.

ALB Application Load Balancer (ALB) instances
OSS Object Storage Service (OSS) buckets
ENI

Only basic GA instances support this type.

Elastic network interfaces (ENIs)

Only secondary ENIs are supported.

Outside Alibaba Cloud Internet Custom IP addresses Custom IP addresses of origin servers
Custom domain names Custom domain names of origin servers
Note
  • By default, you cannot specify ECS instances, internal-facing CLB instances, and internal-facing ALB instances as endpoints for standard GA instances. If you want to specify them as endpoints, submit a ticket.
  • The IP addresses of endpoint groups associated with each GA instance must be globally unique and not conflict with those of other GA instances.
  • You can add only one endpoint to the endpoint group of a basic GA instance. The endpoint must be a CLB instance that is deployed in a VPC or a secondary ENI.

You can specify a weight for an endpoint. The weight specifies the proportion of traffic that is forwarded to the endpoint. GA calculates the sum of all endpoint weights in an endpoint group. Then, traffic is forwarded to each endpoint based on the proportion of each endpoint. For more information, see Set the weight of an endpoint.

Health checks

You can enable health checks for endpoint groups of a GA instance. This improves service reliability and availability and prevents service interruptions caused by unhealthy endpoints.

After you enable health checks for an endpoint group, GA periodically checks whether the endpoints are healthy. When GA detects an unhealthy endpoint, GA distributes new requests to other healthy endpoints. When the unhealthy endpoint recovers, GA distributes requests to the endpoint again. For more information, see Enable and manage health checks

Note You can enable health checks only for endpoint groups of standard GA instances.

References