CheckOps home

Tools directory

C

CheckOps

chaoskube

chaoskube is an MIT tool that periodically kills random pods in a Kubernetes cluster, with filters by namespace, labels, annotations and time windows.

Visit website

Latest public signals

Not refreshed yet. Refresh reads the public site and, when the entry has a GitHub repository, the latest release.

Visit Website

Introduction

chaoskube periodically kills random pods in your Kubernetes cluster. By default it kills a pod in any namespace every ten minutes, and it starts in dry-run mode until you explicitly enable termination.

Filters narrow the targets by namespace, labels, annotations, minimum pod age and pod kinds, and time-based settings exclude weekends, holidays or hours outside working time. It runs as a deployment in the cluster and exposes Prometheus metrics.

In the quality loop chaoskube validates that workloads tolerate random pod loss.

The source code is at linki/chaoskube on GitHub under the MIT licence, with about 1,900 stars; the most recent commit was on 8 October 2026.

Features

  1. Random kills: One pod per interval.
  2. Dry run by default: Safe first runs.
  3. Filters: Namespace, label, annotation and age.
  4. Time windows: Exclude days and hours.
  5. Metrics: Prometheus counters.
  6. Helm chart: Easy deployment.

Pros

  1. Safe defaults: Dry run until enabled.
  2. Free to use: Open source under the MIT licence, so the core tool costs nothing to use.
  3. Actively maintained: The repository had commits as recent as 8 October 2026.

Cons

  1. Pod deletion only: No other fault types.
  2. Not targeted: Random selection.
  3. Minimal reporting: Logs and metrics only.
chaoskube --interval=10m --namespaces='!kube-system' --labels='chaos=enabled' --no-dry-run

Resources:

Pricing

chaoskube is free and open source under the MIT licence.

  1. Teams wanting continuous low-level chaos.

Conclusion

chaoskube is a tidy, configurable random pod killer.

Pair it with alerting to see whether kills cause user impact.

On CheckOps, chaoskube is filed under the Production monitoring & observability stage. Related tools listed here include Chaos Mesh, LitmusChaos, Chaos Toolkit, Gremlin, Steadybit and Toxiproxy. A short trial on a real project, with your own code, data and pipeline, is usually the fastest way to choose between them, because fit matters more than feature lists. Facts here come from the official site, documentation and repository, checked on 11 November 2026.

Alternatives

Other tools on CheckOps that cover similar ground. Open a page, or compare one with chaoskube side by side.