Configuring Scaling Policies

The Cloud Container Engine (CCE) node group supports configuring multiple standby models and can meet resource usage demands across various scenarios when combined with a scaling strategy.

Introduction of scaling strategy

Different scaling strategies correspond to different scaling methods. For specific information, see the description below:

Model configuration sequence

Scale-up principle: Scale-up is performed in the following sequence: active model is selected first; if active model resources are insufficient, standby models are selected successively (in the sequence of addition)

Note: For instance, if you set A as the active model and B and C as standby models, the system will attempt to scale up in the following order: A, B, and then C. If A is unavailable, it will proceed to scale up B, and if B is also unavailable, it will shift to scaling up C.

Scale-down principle: Scale-down occurs in the reverse priority order (the reverse of the node scale-up sequence) defined during scale-up.

Multi-subnet uniform distribution

Scale-up principle: The scale-up is performed in the sequence of models in each subnet to ensure that the count of nodes in each subnet is as equal as possible.

Scale-down principle: Before scale-down, the count of subnets is first calculated. The scale-down is performed according to the subnet node uniform distribution strategy, to ensure the count of nodes in each subnet is as equal as possible.

Prerequisites

• Successfully create K8S Cluster CCE.
  

Operation steps

Specify scaling strategies when creating a NodeGroup

1. Sign in to Cloud Container Engine Console (CCE).
2. Select Cluster List in the left navigation bar.
3. On the Cluster List page, click on the desired cluster to open the cluster management page.
4. Select Node Management > Node Group from the left navigation bar in the cluster management page.
5. On the Node Group page, click Create Node Group to enter the creation page.
6. Configure Scaling Strategy in Advanced Settings.
7. Click the "Finish" button to complete the creation of the node group.
   
   
   

Note:

• This strategy takes effect only when multiple standby models are configured (it is recommended to distribute subnets across different availability zones to enhance the high cross-AZ availability of clusters). For details, refer to Configure Standby Model.

Update the node group scaling strategy

1. Sign in to Cloud Container Engine Console (CCE).
2. Select Cluster List in the left navigation bar.
3. On the Cluster List page, click on the desired cluster to open the cluster management page.
4. Select Node Management > Node Group from the left navigation bar in the cluster management page.
5. On the Node Group page, click the desired node group to open the Node Group Details page.
6. Click Modify in Advanced Configuration.
7. Update the scaling strategy settings in the "Modify Advanced Configuration of Node Group" pop-up window.
8. Click the "OK" button to finalize the update to the scaling strategy.
   
   

Note:

• Modifying the advanced configuration of node groups applies only to new scale-up nodes and does not impact existing nodes or ongoing service operations in the node group.