Manage alarm-triggered policies

Auto Scaling can dynamically adjust the number of KEC instances in a scaling group based on metrics of KEC instances. You can create an alarm-triggered policy to specify a metric, the status of the metric to trigger a scaling activity, and the scaling activity content.

To create an alarm-triggered policy, you need to specify the condition and action.

To define a condition, you need to specify the following information: metric, threshold, cycle, and number of consecutive cycles in which the threshold is reached. The condition is met when the value of the metric reaches the threshold in N consecutive cycles.

The following metrics can be used:

• CPU usage
• Memory usage
• ENI outbound traffic
• ENI inbound traffic
• Listener outbound traffic
• Listener inbound traffic

The action to be taken by Auto Scaling is to send a notification and add or remove the specified number of KEC instances.

Create an alarm-triggered policy

1. Log in to the Auto Scaling console.

2. Click Scaling Group in the left navigation pane.

3. Select a region from the top navigation bar.

4. In the scaling group list, find the scaling group for which you want to create an alarm-triggered policy, and click Manage in the Operation column. The details page of the scaling group appears.

5. Click the Alarm-Triggered Policy tab. The list of alarm-triggered policies for the scaling group appears.

6. Click Create.

7. In the Create Alarm-Triggered Policy dialog box that appears, set parameters and click OK.

Delete an alarm-triggered policy

1. In the scaling group list, find the scaling group for which you want to delete an alarm-triggered policy, and click Manage in the Operation column. The details page of the scaling group appears.

2. Click the Alarm-Triggered Policy tab. The list of alarm-triggered policies for the scaling group appears.

3. Find the alarm-triggered policy to be deleted, and click Delete in the Operation column.

4. In the dialog box that appears, click OK.