kube-monkey is a chaos engineering tool that implements Netflix's Chaos Monkey for Kubernetes clusters.
The tool addresses the need to validate that services can survive unexpected failures by randomly terminating pods within a cluster. Applications opt in through Kubernetes labels, and operators define windows of time during which terminations should occur. Dry run mode is enabled by default, allowing teams to observe the tool's behavior before it actually deletes pods in production environments.
Teams should adopt kube-monkey if they want to systematically test the resilience of their Kubernetes workloads without manual intervention. It suits any organization running stateless or properly replicated services that should tolerate pod loss. The tool is particularly valuable for catching configuration issues where services lack proper health checks or replica counts. Adoption requires only adding a label to the apps you want to test and configuring the schedule and kill modes through the documentation.
The project maintains active documentation covering installation, configuration, and operational concerns including Prometheus metrics and Slack notifications. Development activity shows consistent attention to the codebase with regular updates addressing both new features and maintenance needs. The project accepts community contributions and maintains a dedicated communication channel for users and developers.
