CronHPA Policies

Prerequisites

The CCE Advanced HPA add-on has been installed in the cluster.

Using CronHPA to Adjust the HPA Scaling Scope

CronHPA can periodically scale out/in pods in HPA policies to satisfy complex services.

HPA and CronHPA associate scaling objects using the scaleTargetRef field. If a Deployment is the scaling object for both CronHPA and HPA, the two scaling policies are independent of each other. The operation performed later overwrites the operation performed earlier. As a result, the scaling effect does not meet the expectation.

When CronHPA and HPA are used together, CronHPA rules take effect based on the HPA policy. CronHPA uses HPA to perform operations on the Deployment. Understanding the following parameters can better understand the working rules of the CronHPA.

• targetReplicas: Number of pods set for CronHPA. When CronHPA takes effect, this parameter adjusts the maximum or minimum number of pods in HPA policies to adjust the number of Deployment pods.
• minReplicas: Minimum number of Deployment pods.
• maxReplicas: Maximum number of Deployment pods.
• replicas: Number of pods in a Deployment before the CronHPA policy takes effect.

When the CronHPA rule takes effect, the maximum or minimum number of pods are adjusted by comparing the number of targetReplicas with the actual number of pods and combining the minimum or maximum number of pods in the HPA policy.

Figure 1 CronHPA scaling scenarios

Figure 1 shows possible scaling scenarios. The following examples detail how CronHPA modifies the number of pods in HPAs.

Using the CCE console

1. Log in to the CCE console and click the cluster name to access the cluster console.
2. Choose Workloads in the navigation pane. Locate the target workload and choose More > Auto Scaling in the Operation column.
   Figure 2 Scaling a workload
   

3. Set Policy Type to HPA+CronHPA and enable HPA and CronHPA policies.

   CronHPA periodically adjusts the maximum and minimum numbers of pods using the HPA policy.

4. Configure the HPA policy. For details, see HPA Policies.
   Figure 3 Enabling the HPA policy
   

   Table 2 HPA policy parameters

   Parameter	Description
   Pod Range	Minimum and maximum numbers of pods. When a policy is triggered, the pods are scaled within this range.
   Scaling Behavior	Default: Scale workloads using the Kubernetes default behavior. For details, see Default Behavior. Custom: Scale workloads using custom policies such as stabilization window, steps, and priorities. Unspecified parameters use the values recommended by Kubernetes. Disable scale-out/scale-in: Select whether to disable scale-out or scale-in. Stabilization Window: a period during which CCE continuously checks whether the metrics used for scaling keep fluctuating. CCE triggers scaling if the desired state is not maintained for the entire window. This window restricts the unwanted flapping of pod count due to metric changes. Step: specifies the scaling step. You can set the number or percentage of pods to be scaled in or out within a specified period. If there are multiple policies, you can select the policy that maximizes or minimizes the number of pods.
   System Policy	Metric: You can select CPU usage or Memory usage. NOTE: Usage = CPUs or memory used by pods/Requested CPUs or memory Desired Value: Enter the desired average resource usage. This parameter indicates the desired value of the selected metric. Number of pods to be scaled (rounded up) = (Current metric value/Desired value) x Number of current pods NOTE: When calculating the number of pods to be added or reduced, the HPA policy uses the maximum number of pods in the last 5 minutes. Tolerance Range: Scaling is not triggered when the metric value is within the tolerance range. The desired value must be within the tolerance range. If the metric value is greater than the scale-in threshold and less than the scale-out threshold, no scaling is triggered. This parameter is only supported in clusters v1.15 or later.

5. Click in the CronHPA policy rule. In the dialog box displayed, configure scaling policy parameters.
   Figure 4 Enabling the CronHPA policy
   

   Table 3 CronHPA policy parameters

   Parameter	Description
   Target Instances	When the policy is triggered, CCE will adjust the number of HPA policy pods based on service requirements. For details, see Table 1.
   Trigger Time	You can select a specific time every day, every week, every month, or every year. NOTE: This time indicates the local time of where the node is deployed.
   Enable	Enable or disable the policy rule.

6. After configuring the preceding parameters, click OK. Then, the added policy rule is displayed in the rule list. Repeat the preceding steps to add multiple policy rules, but the triggering time of the policies must be different.
7. Click Create.

Using the kubectl command

When the CronHPA is compatible with the HPA policy, the scaleTargetRef field in CronHPA must be set to the HPA policy, and the scaleTargetRef field in the HPA policy must be set to Deployment. In this way, CronHPA adjusts the maximum and minimum numbers of pods in the HPA policy at a fixed time and the scheduled scaling is compatible with the auto scaling.

1. Create an HPA policy for the Deployment.

   apiVersion: autoscaling/v1
   kind: HorizontalPodAutoscaler
   metadata:
     name: hpa-test
     namespace: default
   spec:
      maxReplicas: 10               #  Maximum number of pods
     minReplicas: 5               #  Minimum number of pods
     scaleTargetRef:              #  Associate a Deployment.
       apiVersion: apps/v1
       kind: Deployment
       name: nginx
     targetCPUUtilizationPercentage: 50

2. Create a CronHPA policy and associate it with the HPA policy created in 1.

   apiVersion: autoscaling.cce.io/v2alpha1
   kind: CronHorizontalPodAutoscaler
   metadata:
     name: ccetest
     namespace: default
   spec:
      scaleTargetRef:                  # Associate an HPA policy.
         apiVersion: autoscaling/v1
         kind: HorizontalPodAutoscaler
         name: hpa-test
      rules:
      - ruleName: "scale-down"    
        schedule: "15 * * * *"         # Running time and period of a job. For details, see Cron, for example, 0 * * * * or @hourly.
        targetReplicas: 1              # Number of target pods
        disable: false
      - ruleName: "scale-up"
        schedule: "13 * * * *"
        targetReplicas: 11
        disable: false

   Table 4 Key fields of CronHPA

   Field	Description
   apiVersion	API version. The value is fixed at autoscaling.cce.io/v2alpha1.
   kind	API type. The value is fixed at CronHorizontalPodAutoscaler.
   metadata.name	Name of a CronHPA policy.
   metadata.namespace	Namespace to which the CronHPA policy belongs.
   spec.scaleTargetRef	Specifies the scaling object of CronHPA. The following fields can be configured: apiVersion: API version of the CronHPA scaling object. kind: API type of the CronHPA scaling object. name: Name of the CronHPA scaling object. CronHPA supports HPA policies or Deployments. For details, see Using CronHPA to Adjust the HPA Scaling Scope or Using CronHPA to Directly Adjust the Number of Deployment Pods.
   spec.rules	CronHPA policy rule. Multiple rules can be added. The following fields can be configured for each rule: ruleName: CronHPA rule name, which must be unique. schedule: Running time and period of a job. For details, see Cron, for example, 0 * * * * or @hourly. NOTE: This time indicates the local time of where the node is deployed. targetReplicas: indicates the number of pods to be scaled in or out. disable: The value can be true or false. false indicates that the rule takes effect, and true indicates that the rule does not take effect.

Using CronHPA to Directly Adjust the Number of Deployment Pods

CronHPA adjusts the associated Deployments separately to periodically adjust the number of Deployment pods. The method is as follows:

Using the CCE console

1. Log in to the CCE console and click the cluster name to access the cluster console.
2. Choose Workloads in the navigation pane. Locate the target workload and choose More > Auto Scaling in the Operation column.
   Figure 5 Scaling a workload
   

3. Set Policy Type to HPA+CronHPA, disable HPA, and enable CronHPA.

   CronHPA periodically adjusts the number of workload pods.

4. Click in the CronHPA policy rule. In the dialog box displayed, configure scaling policy parameters.
   Figure 6 Using CronHPA to adjust the number of workload pods
   

   Table 5 CronHPA policy parameters

   Parameter	Description
   Target Instances	When a policy is triggered, the number of workload pods will be adjusted to the value of this parameter.
   Trigger Time	You can select a specific time every day, every week, every month, or every year. NOTE: This time indicates the local time of where the node is deployed.
   Enable	Enable or disable the policy rule.

5. After configuring the preceding parameters, click OK. Then, the added policy rule is displayed in the rule list. Repeat the preceding steps to add multiple policy rules, but the triggering time of the policies must be different.
6. Click Create.

Using kubectl commands

apiVersion: autoscaling.cce.io/v2alpha1
kind: CronHorizontalPodAutoscaler
metadata:
  name: ccetest
  namespace: default
spec:
   scaleTargetRef:             # Associate a Deployment.
      apiVersion: apps/v1
      kind: Deployment
      name: nginx
   rules:
   - ruleName: "scale-down"
     schedule: "08 * * * *"    # Running time and period of a job. For details, see Cron, for example, 0 * * * * or @hourly.
     targetReplicas: 1
     disable: false
   - ruleName: "scale-up"
     schedule: "05 * * * *"
     targetReplicas: 3
     disable: false