Introduction

Deletes one pod after the other for a given Kubernetes deployment over a configurable period of time. The delay between pod deletions is {attack duration} / {number of pods in deployment}.

Stopping this attack will cause no further pods to be deleted. Kubernetes is responsible for pod restoration/restarts.

Use Cases

• To test the resiliency of a deployment to pod failures
• Check the behavior of the application when a pod is deleted and the pod is restored by Kubernetes
• Check if your load balancer can handle pod failures

Parameters