Scheduled scaling (CronHPA)

Install the CronHPA controller

You can install the CronHPA controller ack-kubernetes-cronhpa-controller by using one of the following methods.

1. Log on to the ACS console. In the left navigation pane, click Clusters.

2. On the Clusters page, click the name of the target cluster. In the left navigation pane, click Add-ons.

3. On the Add-ons page, click the Manage Applications tab. Find ack-kubernetes-cronhpa-controller, click Install, and then click OK in the dialog box that appears.

Create CronHPA jobs

Before you create and run CronHPA jobs for your application, make sure that the CronHPA controller runs as normal in your cluster and only one HPA task is created for your application. For more information about how to enable CronHPA and HPA to interact without conflicts, see Enable CronHPA and HPA to interact without conflicts. You can create CronHPA jobs by using one of the following methods:

Method 1: Create a job during application creation

Method 2: Create a job for an existing application

The following example demonstrates how to create a CronHPA job for an existing application. A stateless application is used in this example.

1. Log on to the ACS console. In the left navigation pane, click Clusters.

2. On the Clusters page, click the name of the target cluster. In the left navigation pane, choose Workloads > Deployments.

3. On the Deployments page, find the application, and then click Details in the Actions column.

4. Click the Pod Scaling tab and configure the scheduled scaling job.

   • If the CronHPA component is not installed, an Install button appears. Click Install and then proceed to the next step.

   • If the CronHPA controller is installed, perform the following steps.

5. Click Create to the right of CronHPA. In the Create dialog box, set the parameters for the job.

   The following table describes the parameters.

   Parameter	Description
   Job Name	Enter a name for the CronHPA job. The name of each CronHPA job must be unique.
   Desired Number of Replicas	Replicated pods are scaled to the desired number at the scheduled time.
   Scaling Schedule	Set the scaling schedule. For more information about how to set the scaling schedule for a CronHPA job, see predefined-schedules.

Create or modify CronHPA jobs

1. Follow the steps in Create a scheduled scaling CronHPA task to go to the Pod Scaling page.

2. On the Pod Scaling tab, in the CronHPA section, find the target job and click Add or Modify Job under the Actions column.

3. In the Edit dialog box, click Add Task to create a new job or click an existing job to modify it. When you are finished, click OK.

   The job configuration form includes the following fields: Job Name, Desired Number of Replicas, and Scaling Schedule (which supports five methods: By Hour, By Day, By Week, By Month, or cron expression). If you select cron expression, enter the expression in the input box in the format: Seconds Minutes Hours Day-of-month Month Day-of-week (for example, 1 2 3 4 * * means 03:02:01 on the 4th of every month).

   Note

   You can delete a CronHPA job. In the Edit dialog box, existing jobs appear as tags. You can remove a job by clicking the delete icon next to its name. To add a new job, enter a name in the Job Name input box and click Add Job. After making your changes, click OK.

Templates of CronHPA and HPA

Enable CronHPA and HPA to interact without conflicts

CronHPA and HPA take effect on workloads, such as Deployments or StatefulSets. If you use CronHPA and HPA to scale a workload at the same time, conflicts may occur. To resolve the conflicts, CronHPA allows you to use HPA as the object to be scaled, which is compatible with the features of HPA.

If both CronHPA and HPA are deployed, CronHPA and HPA may scale pods for the same application that is specified by scaleTargetRef. CronHPA and HPA are independent and unaware of each other. As a result, the CronHPA controller and the HPA controller separately scale pods for the application. The later scaling activity overwrites the earlier activity.

Compatibility solution for CronHPA and HPA

As mentioned, the conflict can be resolved if CronHPA is aware of the current state of HPA. To achieve this, ACK configures the scaleTargetRef in the CronHPA to point to the HPA object itself, not the final workload object. When performing a scaling operation, CronHPA first finds the real scaleTargetRef through the HPA object and then adjusts the workload object based on this state. This prevents the controllers from overwriting each other and ensures that scaling is consistent and predictable.

apiVersion: autoscaling.alibabacloud.com/v1beta1
kind: CronHorizontalPodAutoscaler
metadata:
  labels:
    controller-tools.k8s.io: "1.0"
  name: cronhpa-sample
spec:
   scaleTargetRef:
      apiVersion: autoscaling/v1
      kind: HorizontalPodAutoscaler
      name:  nginx-deployment-basic-hpa
   jobs:
   - name: "scale-down"
     schedule: "30 */1 * * * *"
     targetSize: 1
     runOnce: false
   - name: "scale-up"
     schedule: "0 */1 * * * *"
     targetSize: 3
     runOnce: false

Based on the CronHPA definition, a CronHPA object can determine the values of spec.minReplicas, spec.maxReplicas, and status.desiredReplicas in the HPA. It also knows the current replica count of the workload specified by spec.scaleTargetRef. A CronHPA interacts with the HPA by adjusting it. The CronHPA uses the greater value between its target replica count and the current replica count to determine whether to scale the workload or increase the HPA's upper limit. It uses the smaller value between its target replica count and the HPA's configuration to determine whether to modify the HPA's lower limit.

CronHPA does not directly change the number of pods for the Deployment. Instead, CronHPA triggers HPA to scale the pods. This resolves the conflict between CronHPA and HPA.