Cluster Autoscaler on Cluster API is an open-source project in the K8s community used to implement automatic scaling for Kubernetes clusters. AKE integrates Cluster Autoscaler on Cluster API and provides it to users as the Cluster Autoscaler cluster addon.
After the addon is enabled, users can easily configure node group autoscaling parameters for VM-based workload clusters through the AKE UI to implement node group autoscaling. This function can monitor CPU and memory resource usage of K8s clusters in real time and dynamically adjust the number of nodes in node groups based on parameters configured by users, effectively managing cluster resources while meeting workload requirements. When cluster nodes do not have sufficient resources to schedule newly created Pods, the number of nodes in the corresponding node group is automatically scaled out to ensure that Pods run properly. When resources of cluster nodes are idle, the number of nodes in the corresponding node group is automatically scaled in, releasing resources occupied by node virtual machines to improve resource utilization.
