Deploying the Kubernetes Metrics Server on a Cluster

Find out how to deploy the Kubernetes Metrics Server as a standalone program or as a cluster add-on, on a cluster you've created using Kubernetes Engine (OKE).

You can deploy the Kubernetes Metrics Server on clusters you create using Kubernetes Engine to enable autoscaling.

The Kubernetes Metrics Server is a cluster-wide aggregator of resource usage data. The Kubernetes Metrics Server collects resource metrics from the kubelet running on each worker node and exposes them in the Kubernetes API server through the Kubernetes Metrics API. Other Kubernetes add-ons require the Kubernetes Metrics Server, including:

• the Horizontal Pod Autoscaler (see Using the Kubernetes Horizontal Pod Autoscaler)
• the Vertical Pod Autoscaler (see Using the Kubernetes Vertical Pod Autoscaler)

You can deploy the Kubernetes Metrics Server on a Kubernetes cluster in two ways:

• as a standalone program, on clusters with managed node pools or virtual node pools (see Working with the Kubernetes Metrics Server as a Standalone Program)
• as a cluster add-on, on clusters with managed node pools (see Working with the Kubernetes Metrics Server as a Cluster Add-on)

Note that the Kubernetes Metrics Server is not intended to be used for anything other than autoscaling. For example, it is not recommended that you use the Kubernetes Metrics Server to forward metrics to monitoring solutions, nor as a source of monitoring solution metrics. For more information, see the Kubernetes Metrics Server documentation on GitHub.