Health check

An SLB instance periodically pings all its backend servers to check their operating status. This operation is called health check.

An SLB instance does not send requests to an unhealthy backend server. After the unhealthy backend server recovers, the SLB instance continues to send requests to the backend server.

Health check parameters

Health Check: Specifies whether to enable health check.

Check Interval (S): The health check interval.

Healthy Threshold (Count): If a backend server passes the specified number of consecutive health checks, the system considers the backend server healthy.

Unhealthy Threshold (Count): If a backend server fails the specified number of consecutive health checks, the system considers the backend server unhealthy.

HTTP Request URL: Available only for the HTTP or HTTPS health check. Health checks will use the HEAD method to request this URL. Make sure that the connection string you specify is correctly returned when the corresponding HEAD method is used.

Domain Name: The service domain name of the health check, which is available only for the HTTP or HTTPS health check. By default, an VIP is used as the host.

Health check configuration for Layer 4 forwarding

An SLB instance initiates an access request to the specified backend server port. If the SLB instance receives a response as expected, it considers that the backend server is operating correctly. If not, the SLB instance considers that the backend server is not operating correctly. For TCP services, SYN packets are used for probing.

• Health Check: Enable health check.
• Check Interval (S): Valid values range from 2 to 300 seconds. The default value is 5 seconds.
• Healthy Threshold (Count): Valid values range from 2 to 10. The default value is 5. If an unhealthy backend server passes the specified number of consecutive health checks, the SLB instance considers that the backend server has recovered.
• Unhealthy Threshold (Count): Valid values range from 2 to 10. The default value is 4. If a healthy backend server fails the specified number of consecutive health checks, the SLB instance considers that the backend server has become unhealthy.

Health check configuration for Layer 7 forwarding

An SLB instance sends HTTP requests to backend servers and checks the returned values to identify the health status of the backend servers.

• Health Check: Enable health check.
• Check Interval (S): Valid values range from 2 to 300 seconds. The default value is 5 seconds.
• Healthy Threshold (Count): Valid values range from 2 to 10. The default value is 5. If an unhealthy backend server passes the specified number of consecutive health checks, the SLB instance considers that the backend server has recovered.
• Unhealthy Threshold (Count): Valid values range from 2 to 10. The default value is 4. If a healthy backend server fails the specified number of consecutive health checks, the SLB instance considers that the backend server has become unhealthy.