Krkn Operator

A Kubernetes-native platform for orchestrating Chaos Engineering experiments through a web interface.

Krkn Operator is a Kubernetes-native platform to deploy, configure and orchestrate Chaos Engineering experiments through a web console. Instead of manually executing scenarios, the Operator provides a complete platform to manage users, clusters, permissions and experiment execution from a single place.


Multi-Cluster Chaos Testing

Multi-cluster execution is a core capability of the platform. Every chaos scenario can target one or more clusters simultaneously, regardless of how those clusters were registered:

  • Manual targets — clusters registered directly by an administrator through the web console
  • OCM/ACM discovery — clusters discovered automatically through Open Cluster Management or Red Hat Advanced Cluster Management

Both methods produce the same unified cluster pool. When executing a scenario, users select any combination of available targets and the platform runs the experiment in parallel across all of them. This enables cross-environment resilience testing (dev, staging, production), regional failover validation, and multi-tenant chaos campaigns — all from a single execution.


Personas

The platform defines two roles with different access levels.

Admin

Full access to the platform. Administrators configure the infrastructure that users operate on.

  • Register and remove target clusters
  • Create groups with cluster access and permissions
  • Create users and assign them to groups
  • Configure private registries and their visibility
  • Configure target providers (e.g. ACM integration)
  • All operational features available to users

User

Operational access scoped by group membership. Every user must belong to a group — the group determines which clusters, registries and features are accessible.

  • View and manage jobs (based on View/Delete permissions)
  • Use the cluster terminal (requires Run permission)
  • Run scenarios on assigned clusters (requires Run permission)
  • Design workflows in Chaos Studio (requires Run permission)
  • Upload and manage files (scoped by group visibility)

Permission Model

        Admin
    Creates Groups
    ┌─────┴──────┐
    │             │
  Group A      Group B
    │             │
  Clusters     Clusters
  Permissions  Permissions
  (View,Run,   (View)
   Delete)
    │             │
  Users        Users

Permissions are assigned at the group level and determine what users can do on clusters accessible to their group.

View

View

Allows users to see jobs and their execution results. Users with this permission can inspect single run outcomes, console logs, graph run nodes, and Resiliency Scores — but cannot execute or delete anything.

Applies to: Jobs

Run

Run

Allows users to execute chaos experiments on the clusters assigned to their group. This includes running single scenarios, designing and executing Chaos Studio workflows, and accessing the cluster terminal for read-only inspection.

Applies to: Run Scenarios · Chaos Studio · Cluster Terminal

Delete

Delete

Allows users to remove completed jobs and their execution history. Without this permission, jobs remain visible but cannot be deleted.

Applies to: Jobs


Key Capabilities

CapabilityDescription
Cluster ManagementRegister and manage target Kubernetes clusters
User & Group ManagementOrganize users through groups with granular permissions
Private RegistriesConfigure private container registries with group-based visibility
Chaos StudioDesign reusable visual workflows with serial and parallel execution
Multi-cluster ExecutionRun experiments on one or more clusters simultaneously
Resiliency ScoreMeasure application resilience using PromQL-based metrics
JobsMonitor experiment progress and inspect execution logs
Cluster TerminalExplore managed clusters using read-only kubectl and oc commands
File ManagementStore reusable configuration files and PromQL queries
Provider ConfigurationConfigure target providers (ACM/OCM integration)

Compatibility

  • Kubernetes
  • OpenShift
  • Open Cluster Management (OCM)
  • Red Hat Advanced Cluster Management (ACM)

See the Compatibility Matrix for tested ACM/OCM versions and platform details.


Next Steps


Installation

Install Krkn Operator using Helm

Administration

Platform administration for Krkn Operator

Usage

Operational features for running chaos experiments

Releases & Roadmap

Krkn Operator release strategy, roadmap, and compatibility matrix

Feedback

Share feedback on the Krkn Operator Developer Preview

OCM/ACM Compatibility

Supported platforms and ACM/OCM versions for Krkn Operator