---
title: "Metrics"
description: "Cluster resource metrics for Kubernetes: what the tab will show, and how to get CPU, memory and dashboards today with add-ons."
category: "Deployment"
language: en
source: https://cubepath.com/docs/products/deployment/k8s-metrics
updated: 2026-08-04
---

# Metrics

Open the cluster → **Metrics** tab in [my.cubepath.com](https://my.cubepath.com).

> Note: This tab is **coming soon**. CPU, memory and network metrics for your cluster nodes will appear here.

## How to get metrics today

Kubernetes ships with no metrics pipeline, so nothing collects usage until you install something. Two options from the [Addons](/docs/products/deployment/k8s-addons) tab:

| Add-on | What it gives you |
|--------|-------------------|
| **Metrics Server** | The basics: `kubectl top nodes`, `kubectl top pods`, and the data the **Horizontal Pod Autoscaler** needs to scale on CPU or memory. No history, no dashboards |
| **Prometheus Stack** | Full monitoring: metric collection with history, alerting rules, and Grafana dashboards |

> Important: Install **Metrics Server** even if you don't want dashboards. Without it `kubectl top` returns an error and the Horizontal Pod Autoscaler can't scale at all — it's a dependency, not a nicety.

## Node-level metrics in the meantime

Each worker is a [VPS](/docs/products/deployment/deploy-a-virtual-machine), so its own [Metrics](/docs/products/deployment/vps-metrics) tab already charts CPU, memory, disk and network for that machine — and you can put a [Cloud Alert](/docs/products/deployment/cloud-alerts) on any of it.

That's the right layer for "is this node saturated?". It won't tell you which pod is responsible, which is exactly what Metrics Server or Prometheus adds.