IP as a Backend

IP as a Backend allows your load balancer to distribute incoming traffic to cloud and on-premises servers.

If you need to add IP addresses of on-premises servers or servers in other VPCs as backend servers, you can IP as a Backend.

Constraints

IP as a backend cannot be disabled after it is enabled.
More IP addresses are required for creating a load balancer with IP as a backend enabled. There is an algorithm to determine how many IP addresses are required.
You can also call APIs to calculate the number of required IP addresses by referring to the Elastic Load Balance API Reference.
Only private IPv4 addresses can be added as backend servers.
A maximum of 100,000 concurrent connections can be established with a backend server that is added by using its IP address.
If you add IP addresses as backend servers, the source IP addresses of the clients cannot be passed to these servers.

Adding IP as Backend Servers

After you enable IP as a Backend, you can add backend servers by using their IP addresses. You need to prepare for different scenarios as shown in Table 1.

**Table 1** Adding IP addresses as backend servers
Where IP Addresses Are Located	Preparations
In a different VPC from the load balancer	Set up a VPC peering connection between the VPC where the load balancer is running and the VPC where the IP addresses are located. For details, see the Virtual Private Cloud User Guide.
In the same VPC as the load balancer	Set up a VPC peering connection to connect the resources in the VPC where the load balancer is located and, and then add routes for the VPC peering connection. For details, see Routing Traffic to Backend Servers in the Same VPC as the Load Balancer.
In on-premises data centers	Connect the on-premises data center to the VPC where the load balancer is running through Direct Connect or VPN. For details about how to connect on-premises data centers to the cloud, see the or Direct Connect User Guide or Virtual Private Network User Guide.