If you are using AS for the first time, follow the wizard-based process to create an AS group. This section introduces how to create an AS group.

Precondition

The required VPC, security group and ELB are created.
If you use certificate authentication, the SSH key is prepared.

Authentication mode applies to ECSs added in elastic scaling activities. Other ECSs in the AS group are not affected.

Operation Steps

Register a eSurfing Cloud account and log in to eSurfing Cloud at https://www.esurfingcloud.com.
Select the Console, switch to the node where you want to create AS, and select AS .
On the AS page, click Create AS Group .
On the page for creating an AS group, enter basic information such as the AS group name, maximum/minimum number of instances, VPC, subnet and LB.

Parameter Descriptions

AS Group name	Name of created AS group
Maximum/minimum number of instances	It indicates the maximum or minimum number of ECSs in an AS group. If the minimum value is not 0, a scaling activity is triggered after an AS group is created. Then, the number of instances in the AS group will be changed to the minimum value.
VPC	ECSs use networks provided by a virtual private cloud (VPC). All ECSs in the same AS group belong to this VPC.
Subnet	By default, all ECSs in a VPC subnet can communicate with one another. You can select multiple subnets for an AS group.
LB	It is an optional parameter. After the load balancer is used, ELB automatically distributes the access traffic to all ECSs in the AS group to extend the service capabilities of the application system and guarantee higher levels of application fault tolerance. After an AS group is associated with a load balance instance, the instance is automatically added to the backend server group regardless of whether the instance is automatically created by the AS group or manually added by the user. Note:An AS group can be added with a maximum of 10 groups of LB listeners. After multiple LB listeners are added, multiple services can be concurrently listened to, thereby improving service scalability. IfELB Health Checkis selected, when any one of the listeners detects that an ECS becomes unhealthy, the AS group will replace it with a new one. If you select ELB, configure the following four parameters: Load balancer, backend ECS group, backend port, and weight.
Instance removal policy	Policy for the priority of instances to be removed. When an AS group is scaling down, the instance removal policy is triggered by one of the following four priorities: l Earlier created configurations and instances: Filter out the instances created by the earlier created configuration first, then filter out the earlier created instances to remove. l Earlier created instances: Filter out the instances created at earlier time. l Later created configurations and instances: Filter out the instances created by a later configuration, and then filter out the instances created later to remove. l Later created instances: Filter out the instances created later.
Pay-per-use instance recycling mode	Processing policy for pay-per-use instances after removal. o Release mode: Release the removed pay-per-use ECSs. o Shutdown recycle mode: Shut down but do not release the removed ECSs. These ECSs will be re-added to the AS group preferentially if expansion needed. Note: If ECS with annual or monthly package is removed from the AS group, resources will not be released and states will not be changed. This policy does not apply.
Health check methods	Health check removes the abnormal ECS from the AS group and re-creates a new ECS. You can use the following two health check methods for AS groups: o ECS health check: It is a check of the operation states of ECS. Shutdown and deleted are abnormal states of ECS. This option is selected by default. Using this option, the AS group periodically evaluates the running status of each ECS based on the health check results. If the health check results show that an ECS is unhealthy, AS removes the ECS from the AS group and re-creates an ECS. o ELB health check: It is a check of ECS based on the health check results of ELB. If an AS group uses an elastic load balancer, you can use this option to perform a health check. ECSs detected by all listeners must be normal. Otherwise, AS removes the ECS from the AS group and re-creates one.
Health check interval	It specifies the period between health checks. You can set a health check interval, such as 5 minutes, 15 minutes, 1 hour, or 3 hours, based on the service requirements and to ensure that it includes the expected startup time of the application in the ECS.

Click Next . The page for creating AS configuration items is displayed.

On the displayed page, you can use an existing AS configuration item or create a new AS configuration item. You can modify the AS configuration for the AS group after it is created. To create an AS configuration item, enter AS configuration information, such as configuration name, ECS image, ECS type and login mode.

Auto Scaling

Creating an AS Group Rapidly