Getting started with a custom routing accelerator - AWS Global Accelerator

Getting started with a custom routing accelerator

This section provides steps for creating a custom routing accelerator, which routes traffic deterministically to Amazon EC2 instance destinations in virtual private cloud (VPC) subnet endpoints.


Before you begin

Before you create a custom routing accelerator, create a resource that you can add as an endpoint to direct traffic to. A custom routing accelerator endpoint must be a virtual private cloud (VPC) subnet, which can include multiple Amazon EC2 instances. For instructions for creating the resources see the following:

When you create a resource to add to Global Accelerator, be aware of the following:

  • When you add an EC2 instance endpoint in Global Accelerator, you enable internet traffic to flow directly to and from the endpoint in VPCs by targeting it in a private subnet. The VPC that contains the EC2 instance must have an internet gateway attached to it, to indicate that the VPC accepts internet traffic. For more information, see Secure VPC connections in AWS Global Accelerator.

Before you create a custom routing accelerator, make sure that you review the best practices described in Guidelines and restrictions for custom routing accelerators.

Step 1: Create a custom routing accelerator

To create an accelerator

  1. Open the Global Accelerator console at

  2. Provide a name for your accelerator.

  3. For Accelerator type, select Custom routing.

  4. Optionally, add one or more tags to help you identify your accelerator resources.

  5. Choose Next to add listeners, endpoint groups, and VPC subnet endpoints.

Step 2: Add listeners

Create a listener to process inbound connections from your users to Global Accelerator.

The range that you specify when you create a listener defines how many listener port and destination IP address combinations that you can use with your custom routing accelerator. For maximum flexibility, we recommend that you specify a large port range. Each listener port range that you specify must include a minimum of 16 ports.

To create a listener

  1. On the Add listener page, enter the ports or port ranges that you want to associate with the listener. Listeners support ports 1-65535.

  2. Choose the protocol or protocols for the ports that you entered.

  3. Optionally, choose Add listener to add an additional listener.

  4. When you're finished adding listeners, choose Next.

Step 3: Add endpoint groups

Add one or more endpoint groups, each of which is associated with a specific AWS Region. For each endpoint group, specify one or more sets of port ranges and protocols. Global Accelerator uses these to direct traffic to Amazon EC2 instances in subnets in the Region.

For each port range that you provide, you also specify the protocol to use: UDP, TCP, or both UDP and TCP.

To add an endpoint group

  1. On the Add endpoint groups page, in the section for a listener, choose a Region.

  2. For Ports and protocols sets, enter port ranges and protocols for your Amazon EC2 instances.

    • Enter a From port and a To port to specify a range of ports.

    • For each port range, specify the protocol or protocols for that range.

    The port range doesn't have to be a subset of your listener port range, but there must be enough total ports in the listener port range to support the total number of ports that you specify.

  3. Choose Save.

  4. Optionally, choose Add endpoint group to add additional endpoint groups for this listener or other listeners.

  5. Choose Next.

Step 4: Add VPC subnet endpoints

Add one or more virtual private cloud (VPC) subnet endpoints for this regional endpoint group. Endpoints for custom routing accelerators define the VPC subnets that can receive traffic through a custom routing accelerator. Each subnet can contain one or many Amazon EC2 instance destinations.

When you add a VPC subnet endpoint, Global Accelerator generates new port mappings that you can use to route traffic to the destination EC2 instance IP addresses in the subnet. Then you can use the Global Accelerator API to get a static list of all the port mappings for the subnet, and use the mapping to deterministically direct traffic to specific EC2 instances.

To add endpoints

  1. On the Add endpoints page, in the section for the endpoint group that you want to add the endpoint to, choose a subnet ID for Endpoint.

  2. Optionally, do one of the following to enable traffic to EC2 instance destinations in the subnet:

    • To allow traffic to be directed to all EC2 endpoints and ports on the subnet, select Allow all traffic

    • To allow traffic to specific EC2 endpoints and ports on the subnet, select Allow traffic to specific destination socket addresses. Then specify the IP addresses and ports or port ranges to allow. Finally, choose Allow these destinations.

    By default, no traffic is allowed to subnet endpoints. If you don't select an option to allow traffic, traffic is denied to all destinations in the subnet.


    If you want to enable traffic to specific EC2 instances and ports in the subnet, you can do that programmatically. For more information, see AllowCustomRoutingTraffic in the AWS Global Accelerator API Reference.

  3. Choose Next.

After you choose Next, on the Global Accelerator, dashboard you'll see a message that your accelerator is in progress. When the process is finished, the accelerator status in the dashboard is Active.

Step 5 (optional): Delete your accelerator

If you created an accelerator as a test or if you're no longer using an accelerator, you can delete it. On the console, disable the accelerator, and then you can delete it. You don't have to remove listeners and endpoint groups from the accelerator.

To delete an accelerator by using an API operation instead of the console, you must first remove all listeners and endpoint groups that are associated with the accelerator as well as disable it. For more information, see the DeleteCustomRoutingAccelerator operation in the AWS Global Accelerator API Reference.

Be aware of the following when you delete an accelerator:

  • When you create an accelerator, Global Accelerator provides you with a set of two static IP addresses. The IP addresses are assigned to your accelerator for as long as it exists, even if you disable the accelerator and it no longer accepts or routes traffic. However, when you delete an accelerator, you lose the static IP addresses that are assigned to the accelerator, so you can no longer route traffic by using them. As a best practice, ensure that you have permissions in place to avoid inadvertently deleting accelerators. You can use IAM policies like tag-based permissions with Global Accelerator to limit the users who have permissions to delete an accelerator. For more information, see ABAC with Global Accelerator.

To delete an accelerator

  1. Open the Global Accelerator console at

  2. Choose the accelerator that you want to delete.

  3. Choose Edit.

  4. Choose Disable accelerator, and then choose Save.

  5. Choose the accelerator that you want to delete.

  6. Choose Delete accelerator.

  7. In the confirmation dialog box, choose Delete.