kube-monkey
github.com
- Free plan
- Self-hosted
What kube-monkey is
kube-monkey is a self-hosted Kubernetes tool that randomly terminates pods to help teams develop and validate services that can withstand failures. Applications opt in using Kubernetes labels, including an enabled setting and values such as an identifier and MTBF. Dry run is on by default, so planned terminations are logged rather than executed. You can set run days, schedule creation time, termination hours, timezone, and pod shutdown grace period. Deployments, StatefulSets, and DaemonSets are supported directly; other resource kinds can be configured as custom resources with matching RBAC. Namespace whitelists and blacklists help limit eligible workloads, with kube-system excluded by default. The project documents Slack or user-provided API notifications and a Prometheus metrics endpoint. Helm installation instructions are provided for deploying it into kube-system. Namespace pattern filtering requires cluster-wide workload list permission; a namespace-scoped Role installation logs an error and schedules nothing. kube-monkey is free and distributed under the Apache License 2.0.
Who it suits
It suits Kubernetes teams that want to exercise failure resilience by terminating opted-in pods. Teams should be prepared to manage its cluster permissions and scheduling controls.
Good to know, and what to watch
- Dry run logs planned terminations by default.
- Scheduling includes days, hours, timezone, and grace period.
- Namespace lists constrain eligible workloads.
- Documents Slack or API notifications and Prometheus metrics.
- Namespace filtering requires cluster-wide workload list permission.
- A namespace-scoped Role logs an error and schedules nothing.
Plans
Every plan and pricekube-monkey
Open source
The verdict
kube-monkey provides configurable pod termination for Kubernetes failure-resilience work, with a default dry run and namespace controls. Check the required cluster-wide permission before relying on namespace pattern filtering.
What the maker publishes
- Installation
- The README provides Helm installation instructions for deploying kube-monkey into the kube-system namespace. source
- License
- The repository is distributed under the Apache License 2.0. source
- Metrics
- The project documentation lists a Prometheus metrics endpoint. source
- Namespace controls
- A whitelist can restrict eligible namespaces, a blacklist can exclude namespaces, and the default blacklist includes kube-system. source
- Notable permission requirement
- Namespace pattern filtering requires cluster-wide workload list permission; a namespace-scoped Role installation will log an error and schedule nothing. source
- Notifications
- The project documentation describes notifications that can post attack events to Slack or a user-provided API. source
- Opt-in
- Applications opt in through Kubernetes labels, including an enabled label and settings such as an identifier and MTBF. source
- Purpose
- Implements Netflix’s Chaos Monkey for Kubernetes by randomly deleting pods to help develop and validate failure-resilient services. source
- Safe default
- Dry run is enabled by default and logs planned terminations instead of carrying them out. source
- Scheduling
- Users can configure run days, schedule creation time, termination hours, timezone, and pod shutdown grace period. source
- Support
- The README directs contributors to the Kubernetes Slack #kube-monkey channel and project contribution guidance. source
- Supported workloads
- The project covers Deployments, StatefulSets, and DaemonSets directly; other resource kinds can be configured as custom resources with matching RBAC. source
Checked against the maker's pages on 9 Oct 2026.