k9s: The Kubernetes Terminal UI

Ulrik Christensen

November 17, 2026

If you manage Kubernetes every day, you know the routine: kubectl get pods, copy a pod name, kubectl describe, kubectl logs, switch namespace, and start over. It works, but it is slow, and it only ever shows you one small slice of a cluster that is constantly changing. k9s fixes that. It is an open-source terminal UI that turns your cluster into a live, navigable dashboard without ever leaving the terminal.

A good management console is half the job; k9s is the tool that made Kubernetes administration feel just as manageable to me.

What is k9s?

k9s is a terminal-based user interface for Kubernetes. It reads the same kubeconfig file as kubectl and talks to the same Kubernetes API, using your existing RBAC permissions. That means there is nothing to deploy inside the cluster, no agents, and no extra ports to open. It works the same way against Amazon EKS, Google GKE, Azure AKS, k3s, or a homelab cluster.

Installing it takes seconds:

# macOS / Linux
brew install k9s

# Windows
winget install k9s

# Start k9s in a specific context and namespace
k9s --context prod-cluster -n web

Why administrators should use k9s

A live overview of the whole cluster

kubectl gives you a snapshot. k9s gives you a live view. Pods, deployments, nodes, and services refresh continuously, and status is color-coded, so a pod stuck in CrashLoopBackOff or Pending stands out immediately. Instead of piecing the picture together from a dozen commands, you see the state of your workloads on one screen.

Faster troubleshooting

The classic debug loop of finding a pod, describing it, reading its logs, and opening a shell becomes a few keystrokes. Highlight a pod and press l for logs, d to describe, s for a shell, or Shift+F to port-forward. No more copying long, generated pod names between commands. During an incident, that saved time matters.

Drill down through relationships

k9s lets you move from a deployment to its pods, and from a pod to its containers, by pressing Enter. The :xray view shows how resources relate to each other as a tree, which makes it far easier to understand what a Helm chart or operator has actually created.

Safe context and namespace switching

Working with several clusters is where mistakes happen. k9s always shows the active context and namespace in the header, and switching is as simple as typing :ctx or :ns. For read-only investigations, start it with k9s --readonly to disable every action that modifies the cluster. This is ideal for giving support staff or junior administrators visibility without risk.

Resource usage at a glance

When the metrics-server is installed, k9s shows CPU and memory usage for pods and nodes, including usage as a percentage of requests and limits. The :pulses view gives a quick health summary of the whole cluster. It is not a replacement for Prometheus and Grafana, but it answers the question “what is happening right now?” in seconds.

Keyboard-driven and customizable

Everything in k9s is driven by the keyboard, and the commands mirror the resource names you already know from kubectl. You can add your own aliases, hotkeys, skins, and plugins, so the tool adapts to the way your team works.

Essential k9s shortcuts

Key
Action
:pod, :deploy, :svc, :node
Jump to a resource view
:ctx / :ns
Switch context / namespace
0
Show all namespaces
/
Filter the current view
Enter
Drill down into the resource
d / y
Describe / view YAML
l
View logs
s
Open a shell in the container
e
Edit the resource
Shift+F
Port-forward
Ctrl+D / Ctrl+K
Delete / kill
?
Show all shortcuts
Esc / :q
Go back / quit

kubectl cheat sheet

k9s does not replace kubectl. You still need it for scripts, CI/CD pipelines and quick one-liners. Here are the commands I use most.

Task
Command
Show current context
kubectl config current-context
List contexts
kubectl config get-contexts
Switch context
kubectl config use-context <name>
Set default namespace
kubectl config set-context --current --namespace=<ns>
List pods in all namespaces
kubectl get pods -A
Pods with node and IP
kubectl get pods -o wide
Describe a resource
kubectl describe pod <pod>
Follow logs
kubectl logs -f <pod> -c <container>
Logs from a crashed container
kubectl logs <pod> --previous
Shell into a pod
kubectl exec -it <pod> -- /bin/sh
Apply a manifest
kubectl apply -f deployment.yaml
Scale a deployment
kubectl scale deploy <name> --replicas=3
Restart a deployment
kubectl rollout restart deploy <name>
Check rollout status
kubectl rollout status deploy <name>
Roll back
kubectl rollout undo deploy <name>
Port-forward
kubectl port-forward svc/<svc> 8080:80
Resource usage
kubectl top pods / kubectl top nodes
Recent events
kubectl get events --sort-by=.lastTimestamp
Drain a node
kubectl drain <node> --ignore-daemonsets
Export YAML
kubectl get deploy <name> -o yaml

k9s and kubectl: better together

Use kubectl for automation and repeatable changes, and k9s for everything you do interactively: exploring, monitoring, and troubleshooting. Together they give administrators both precision and a clear overview of what is running in their clusters. If you work with Kubernetes on AWS, Google Cloud, or in your homelab, install k9s today. It is one of the easiest productivity wins in the Kubernetes ecosystem.