Server Load Balancers

The Server Load Balancer distributes the incoming traffic among multiple ECS instances providing enhanced availability to your solution. This load balancing is based upon preconfigured  forwarding rules. 

SLB  uses probes to check the health status of the added backend servers. It also automatically provisions ECS instances to eliminate single point of failure and improving the overall service capability and availability of your solution. 

Server Load Balancer consists of the following components:

Server Load Balancer instances: a running load balancing service that receives and distributes the incoming traffic to the backend servers.

Listeners: A listener checks the client requests and forwards the requests to the backend servers. It also performs health check on the backend servers.  Up to 50 listeners can be added to a Server Load Balancer instance. Each listener corresponds to an application deployed on the EC.

Backend servers: ECS instances added to a Server Load Balancer instance to process the distributed requests. ECS instances can be registered with the same load balancer using multiple ports.

Benefits

• High availability: Server Load Balancer is designed to work in the full-redundancy mode without SPOF. Server Load Balancer supports local and cross-region disaster tolerance. When Server Load Balancer is used together with DNS, the service availability is up to 99.95%. You can scale your service based on the application load, without interrupting services continuity.
• Scalable: You can increase or decrease the number of backend servers as needed to expand the service capabilities of your applications.
• Cost effectiveness: Compared with the traditional hardware load balancing system, Server Load Balancer reduces the cost by 60%.
• Secure: Combined with Alibaba Cloud Security, Server Load Balancer can defend against up to 5 Gbps DDoS attacks, such as HTTP flood and SYN flood attacks.

Limitations

• Server Load Balancer does not support cross-region deployment. Users must ensure that the region for the ECS instances and the Server Load Balancer instance is the same.
• Server Load Balancer does not limit the operating system used in the ECS instances as long as the applications deployed in the ECS instances are the same, and the data is consistent.

Backend Server Groups

By virtualising IP address, the SLB virtualises the added ECS instances located in the same region into a highly available application service pool.
Two kinds of server groups are supported:

Master-slave server group When the master server works normally, requests are distributed to it; when it is down, requests will be distributed to the slave server (available only for layer 4 listeners)

VServer group When you need to distribute different requests to different backend servers, or you want to configure domain name and URL forwarding rules, use Vserver groups

Forwarding Rules

Listeners support three scheduling algorithms:

Round robin: Requests are distributed across the group of backend ECS servers sequentially. Weighted round robin (WRR): You can set a weight for each backend server. Servers with higher weights receive more requests than those with lower weights.
Weighted least connections (WLC): In addition to the weight set to each backend ECS server, the number of connections to the client is also considered.