Features:

The initial virtual machine will be deployed to join the AS group after the AS group is created.

Precautions:

1. The auto-scaling function only supports tenant VPC scenarios.
2. All nodes within the auto-scaling feature are loaded with vAD virtual services. If access is needed outside the tenant, bind the elastic IP for the vAD virtual services.
3. Auto-scaling VM is based on built-in image deployment. When creating the built-in image, ensure the production configuration, such as business service auto-startup, listening port, and database connection, are complete.
4. The threshold value monitored by the AS policy is the average value of all virtual machines in the AS group.
5. The AS group in the enabled state is not allowed to delete. Click the green button in the status bar to enable/disable it.
6. The AS group that has associated AS configuration and AS policy is not allowed to delete.

Prerequisite

1. Adequate resources quota has been allocated for the tenant.
2. 2. A load balancing (vAD) device has been created for the tenant.

Steps:

Log in to the tenant portal, navigate to Reliability and select Auto Scaling.

Click New to create an AS Group by filling up the following information.

1. Name of the AS group.
2. Max VMs: Maximum number of VMs in an AS group when performing scaling.
3. Min VMs: Minimum number of VMs in an AS group when a scaling action is performed.
4. Expected VMs: Automatic scaling is performed to add the expected number of VMs after the AS group is created or edited.
5. Cooldown Period(s): A scaling cooldown helps you prevent your AS group from launching or terminating additional VMs before the effects of previous activities are visible.
6. Resource Pool: Select the resource pool where the tenant is located.
7. Network Type: Only tenant VPC is supported currently.
8. NIC: Select the tenant subnet to access.
9. App Load Balancer: Select the vAD planned for auto-scaling within the tenant.

Click Node Pool under Load Balancing, and a Select Node Pool window will appear and configure the following.

1. Name of the node pool.
2. Load Balancing Method: Passes each new connection request to the next node in line in a rotating sequential manner.
3. Health Check Method: Configure the method of vAD detection to VMs. Some of it required to be configured in vAD.
4. Health Check Condition: When multiple health check methods are configured, At least, or All checks are passed will be considered valid.
5. Recovery Period: When a failed node recovers, it gives the node time to get ready before returning to service. During this time, no requests will be sent to the node from the client. Range: 0-300. 0 indicates disabled.
6. Warm-Up Period: When a failed node recovers, this gives the node time to warm up. During this time, the traffic to the node will be slowly ramped up to prevent the node from being overwhelmed by more connections than it can handle as it starts up.  Range: 0-300 and 0 indicates disabled.

Click OK to return to the Create AS Group page, look for Virtual Service under Load Balancing, click New, then configure the following.

1. Name of the virtual service.
2. Service Type: Select the service type.
3. IP Address: Input the IP address of the service provided by vAD to the outside network.
4. Port Range: Fill in the port range for the external service provided by vAD.

Create a new AS configuration or use an existing AS configuration, which is the step of deploying a virtual machine using a built-in image.

Click Next to associate an existing or new AS policy.

1. Policy Name.
2. Trigger Condition: Support setting the threshold value for the average CPU or memory usage.
3. Duration: When all the virtual machines in the AS group have reached the trigger condition and exceed the duration, the scaling action will be executed.
4. Scaling Action: Create or delete the VMs. vAD exit mechanism will be used for deletion where it will not schedule new connections to this virtual machine or terminate existing connections. The virtual machine will be deleted when the existing connection is disconnected.

Click OK to begin creating the AS policy and virtual machine.

The AS group has been created under the associated subnet.

Click Edit to view the status of the node pool.

Click the name of the AS group to view the detailed information about the AS group.