Introduction
Last Updated：2020-07-20
Baidu Load Balance (BLB) implements balanced traffic for applications and achieves business-level scaling in performance by virtualizing multiple Baidu AI Cloud servers in the same region into a group, setting an internal or external service address and forwarding frontend concurrent access to multiple backend cloud servers. BLB also eliminates single-point failures of service in a timely manner through automatic failover to improve service availability.
A new application BLB instance is now available. Its deployment architecture when solving a scenario for the user is as follows:
The performance comparison between the general BLB instance and the newly launched application BLB instance is shown in the following table:
| Instance | Description | Difference |
|---|---|---|
| General BLB instance | Provide conventional load balancing feature components featuring light weight, simple configuration, and low price to meet the general needs of users for load balancing. | The backend servers are organized according to the instance dimension, and all backend servers under each listener have the same listening port. Traffic forwarding methods: Listener unified traffic forwarding supports round robin, least connection, and source IP |
| Application BLB instance | Deploy architectures for users' complex applications, especially large website architectures. Constructed using a policy-based network management framework to achieve business-driven traffic load balancing. | Support server groups. You can organize the backend servers according to the listener dimension. All backend servers under each listener can have different listening ports. Traffic forwarding methods: Forwarding by policy supports traffic forwarding based on domain name/URL |