---
title: "Kubernetes Monitoring overview | Grafana Labs"
description: "Full cluster-to-pod visibility for container orchestration"
---

> For a curated documentation index, see [llms.txt](/llms.txt). For the complete documentation index, see [llms-full.txt](/llms-full.txt).

## Full cluster-to-pod visibility for container orchestration

| Level          | What’s monitored                | Key metrics                     |
|----------------|---------------------------------|---------------------------------|
| **Cluster**    | Control plane, etcd, API server | Availability, latency, pressure |
| **Nodes**      | Kubelet, node resources         | CPU, memory, disk per node      |
| **Pods**       | Container resources, restarts   | Requests vs. usage, OOM kills   |
| **Workloads**  | Deployments, StatefulSets       | Replica availability, rollouts  |
| **Networking** | Services, Ingress, DNS          | Request rates, latency, errors  |

## Out-of-the-box capabilities

[Kubernetes Monitoring capabilities showing dashboards, alerts, cost management, and ML predictions](grafana-cloud-kubernetes-monitoring-1.svg "Kubernetes Monitoring capabilities showing dashboards, alerts, cost management, and ML predictions")

## Questions answered

| With Kubernetes Monitoring, you can answer…                     |
|-----------------------------------------------------------------|
| Which pods are using the most CPU in my cluster?                |
| Why did this pod crash? What were the logs before the OOM kill? |
| How much is my infrastructure costing, and where can I save?    |
| What will my resource usage look like next week?                |
| Which namespace is consuming the most resources?                |

## Problems solved

| Problem                          | Solution                                            |
|----------------------------------|-----------------------------------------------------|
| Complex manual setup             | Helm chart deployment, no manual config             |
| No visibility into pod resources | Complete cluster-to-container metrics               |
| Hard to debug pod crashes        | Correlated logs, metrics, and automated diagnostics |
| Unpredictable costs              | Built-in cost views and savings recommendations     |
| Capacity planning guesswork      | ML-powered CPU and memory predictions               |