LinuxCommandLibrary

ktop

Top-like monitoring tool for Kubernetes clusters

TLDR

Monitor the current Kubernetes cluster
$ ktop
copy
Monitor a specific namespace
$ ktop -n [namespace]
copy
Use a specific kubeconfig
$ ktop --kubeconfig [path/to/config]
copy

SYNOPSIS

ktop [options]

DESCRIPTION

ktop is a terminal UI monitoring tool for Kubernetes clusters, similar to htop for Linux. It provides real-time visualization of cluster resource usage including CPU, memory, disk, and GPU across nodes and pods.
The TUI displays a cluster overview with total resources and aggregate utilization, per-node metrics with CPU, memory, and pod counts, and a sortable list of pods with resource consumption and restart counts. Metrics update every 2 seconds by default.

CAVEATS

Requires a working kubectl configuration and metrics-server deployed in the cluster. GPU monitoring requires appropriate drivers and device plugins.

HISTORY

ktop was created by Vladimir Vivien (vladimirvivien) and is written in Go. It was inspired by classic Unix monitoring tools like top and htop, adapted for Kubernetes cluster monitoring.

SEE ALSO

kubectl(1), k9s(1), htop(1)

> TERMINAL_GEAR

Curated for the Linux community

Copied to clipboard