Term	Description
ELB	It refers to the network load balance service provided by eSurfing Cloud, which consists of four-layer and seven-layer LB services.
Load Balance Instance	It refers to a running load balance service. A load balancing instance must be created before provisioning of LB services.
LB listener	An LB listener monitors the requests on the load balancer, and specifies how to forward the requests to a back-end cloud server for processing.
Backend Server Group	It refers to a set of ECS instances for handling front-end requests distributed by LB.
Backend Server	It refers to ECS instances used to process front-end requests.
Elastic WAN Bandwidth	It refers to the traffic used to provide external services and access the network and through which other network computers can access the server.
Load Balancer Protocol/Port	It supports Layer 4 TCP and Layer 7 HTTP protocols. A port can be selected from ports 1-65535. However, early filing is required to use ports 80, 8080, 443 and 8443. These four filing ports are disabled by default, and will be enabled after filing is completed.
ECS Protocol/Port	It supports Layer 4 TCP and Layer 7 HTTP protocols and specifies the protocol and port of the ECS controlled by LB. A port can be selected from ports 1-65535. ECS ports do not require filing.
Session Persistence	Session persistence can be enabled or disabled. If this function is enabled, cookie-based session persistence is provided for Layer 7 (HTTP) services and IP address-based session persistence is provided for Layer 4 (TCP) services.
Health Check	Users can specify a health check method and frequency. LB regularly checks whether the backend ECS runs properly based on the preset health check rules. Once it detects that ECS is unhealthy, access traffic will not be distributed to the instances on the unhealthy ECS.
Load Mode	It refers to load balancing algorithm and supports three algorithms including polling, minimum number of connections and source address.
Polling	The polling algorithm delivers the requests to the back-end ECS in turn based on the weight of the back-end server and is often used for short-connection services such as HTTP.
Least Connections	This algorithm preferentially sends requests to the backend ECS with the smallest number of connections and is often used for long-connection services such as database connection.
Source Address	The algorithm hashes the source address of the request, and distributes the request to the corresponding ECS based on the weight of the backend ECS. In this way, the request of the same client IP address will be distributed to the specific ECS. It is suitable for LB over TCP protocol without the cookie function.
Default Quota	It refers to the number of resources owned by each user in each regional node. If the default quota cannot meet their requirements, the users can modify the quota on the Console.

Elastic Load Balancing

Basic Concept