Pod Scenarios using krknctl

Scenario-specific krknctl usage for pod disruption runs

Overview

This page summarizes the krknctl form of the Pod Scenarios run command.

For the full pod scenario overview, see the Pod Scenarios page.

krknctl run pod-scenarios [--<parameter>:<value>]

Use this command to run a pod disruption scenario. Pass the optional flags below to target specific pods and control how the run behaves. Global krknctl flags are listed here.

ParameterDescriptionTypeRequiredDefault
--namespaceNamespace to target. Supports regex.stringNoopenshift-*
--pod-labelLabel selector for the pods to target, for example app=test.stringNo-
--exclude-labelPods matching this label are excluded from the run, even if they match other criteria.stringNo""
--name-patternRegex pattern to match pods in the namespace when --pod-label is not set.stringNo.*
--disruption-countNumber of pods to disrupt.numberNo1
--kill-timeoutSeconds to wait for the target pod(s) to be removed.numberNo180
--expected-recovery-timeSeconds to wait for disrupted pods to recover before the scenario fails.numberNo120
--node-label-selectorLabel selector for the nodes that host the target pods.stringNo""
--node-namesExplicit node names to target. Example: ["worker-node-1","worker-node-2"].stringNo[]

Behavior Notes

  • Pods are selected from the namespace and pod filters first, then filtered by node selectors.
  • The scenario fails if the disrupted pods do not recover within --expected-recovery-time seconds.

To see the full CLI help:

krknctl run pod-scenarios --help