Function Overview

Dedicated load balancer is newly launched by Huawei Cloud. It has exclusive use of underlying resources, so that the performance of a dedicated load balancer is not affected by other load balancers. In addition, there are a wide range of specifications available for you to select.

You can create a dedicated load balancer in multiple AZs for higher reliability.

You can add servers in a peer VPC, in a VPC that is in another region and connected through a cloud connection, or in an on-premises data center at the other end of a Direct Connect or VPN connection, by using the server IP addresses. In this way, incoming traffic can be flexibly distributed to cloud servers and on-premises servers for hybrid load balancing.

Note the following when you enable hybrid load balancing:

- Enable the IP as a Backend option when you create a dedicated load balancer. If you want to enable this option for an existing dedicated load balancer, go to its Basic Information page.

- You cannot disable the IP as a Backend option after a dedicated load balancer is created. You can use only IPv4 addresses to add IP as Backend Servers.

- Only TCP, HTTP, and HTTPS listeners support hybrid load balancing. Security group rules of IP as Backend Servers must allow traffic from the backend subnets of the load balancer. Otherwise, health checks will fail.

- Source IP addresses of the clients cannot be passed to IP as Backend Servers. Install the TOA module to obtain source IP addresses.

You can enable slow start for HTTP and HTTPS listeners. After you enable it, the load balancer linearly increases the proportion of requests to backend servers in this mode. When the slow start duration elapses, the load balancer sends full share of requests to backend servers and exits the slow start mode. Slow start ensures that applications start smoothly and can respond to requests with optimal performance. When health checks are enabled, slow start takes effect after backend servers are declared healthy.

Backend servers will exit slow start in either of the following cases:
- The slow start duration elapses.
- The servers become unhealthy during the slow start duration.

After an unhealthy backend server exits slow start, it will once again enter slow start after it is detected as healthy.

Two security policies are available for you to select when you create an HTTPS listener for a dedicated load balancer: TLS-1-0-WITH-1-3 and TLS-1-2-FS-WITH-1-3. The two security policies support TLS 1.3, helping you improve security.

If you add a UDP listener to a dedicated load balancer, you can select Quick UDP Internet Connection (QUIC) for the associated backend server group. In this way, you can use the connection ID algorithm to route requests with the same connection ID to the same backend server. QUIC features low latency and high reliability, does not cause head-of-line blocking, and is ideal for mobile Internet. In addition, no new connections need to be established when you switch between a Wi-Fi and a mobile network.

You can create dedicated load balancers that support either IPv4 or IPv6 or both IPv4 and IPv6.

You can select HTTPS for both your listener and the associated backend server group to verify the validity of requests on your load balancer and backend servers, and to enhance service security. This allows you to enable mutual authentication on the load balancer and backend servers.

When adding a backend server to a dedicated load balancer, you can select a primary NIC or an extension NIC to forward traffic.

You can configure and modify timeout durations (idle timeout, request timeout, and response timeout) for your listeners. For example, if the size of a request from an HTTP or HTTPS client is too large, you can increase the request timeout duration to ensure that the request can be routed successfully.

For shared load balancers, you can change only the timeout durations of TCP, HTTP, and HTTPS listeners, but not UDP listeners. For dedicated load balancers, you can change the timeout durations of TCP, UDP, HTTP, and HTTPS listeners.

Server Name Indication (SNI) is an extension to TLS and is used when a server uses multiple domain names and certificates. You can enable SNI if your applications allow access through multiple domain names and each domain name uses an independent certificate. For example, a single IP address is mapped to multiple domain names.

After you enable SNI, the client can submit the requested domain name at the start of the SSL handshake. After receiving the request, the load balancer searches for the certificate based on the domain name. If the certificate is found, the load balancer will return it to the client. If the certificate is not found, the load balancer will return the default certificate. You can enable SNI only when you add HTTPS listeners. Load balancers can have multiple SNI certificates bound.

Hypertext Transfer Protocol 2.0 (HTTP/2) is a new version of the HTTP protocol. HTTP/2 improves the connections between clients and load balancers. You can enable or disable HTTP/2 for an existing HTTPS listener.

HTTP/2 takes effect only for the connections between the clients and your load balancer. HTTP1.x can be used for connections between your load balancer and backend servers.