Autoscaling Kubernetes Node Pools and Pods

Find out about autoscaling Kubernetes node pools and pods you've created using Kubernetes Engine (OKE).

You can automatically scale the node pools and pods of clusters you create using Kubernetes Engine to optimize resource usage.

To enable cluster autoscaling by autoscaling node pools, you can deploy the Kubernetes Cluster Autoscaler (see Using the Kubernetes Cluster Autoscaler). You can deploy the Kubernetes Cluster Autoscaler on a Kubernetes cluster in two ways:

To enable autoscaling by autoscaling pods, you deploy the Kubernetes Metrics Server to collect resource metrics from each worker node in the cluster (see Deploying the Kubernetes Metrics Server on a Cluster). You can deploy the Kubernetes Metrics Server on a Kubernetes cluster in two ways:

Having deployed the Kubernetes Metrics Server, you can then use:

You can use the Kubernetes Cluster Autoscaler in a cluster with both the Kubernetes Horizontal Pod Autoscaler and the Kubernetes Vertical Pod Autoscaler.

Note

You can use the Kubernetes Cluster Autoscaler, the Kubernetes Metrics Server (as a cluster add-on), and the Kubernetes Vertical Pod Autoscaler with managed node pools only. You can use the Kubernetes Metrics Server (as a standalone program) and the Kubernetes Horizontal Pod Autoscaler with both virtual node pools and managed node pools.