Jump to content

Connect SuperML | Leeroopedia MCP: Equip your AI agents with best practices, code verification, and debugging knowledge. Powered by Leeroo — building Organizational Superintelligence. Contact us at founders@leeroo.com.

Implementation:Kubeflow Pipelines Pipeline Runner Role

From Leeroopedia
Knowledge Sources
Domains RBAC, Kubernetes, Pipeline_Execution
Last Updated 2026-02-13 14:00 GMT

Overview

RBAC Role defining the broad permissions used by the service account that executes pipeline workflow steps at runtime.

Description

The pipeline-runner Role (87 lines) grants the identity under which Argo Workflow pods run pipeline steps. Permissions include: secrets (get), configmaps (get/watch/list), PVs/PVCs (all), volume snapshots (create/delete/get), Argo workflows and workflowtaskresults, pods/exec/logs/services (all), deployments/replicasets (all), all Kubeflow resources, batch jobs (all), and Seldon deployments (all).

Usage

Applied during KFP deployment. The pipeline-runner ServiceAccount is bound to this role and used by Argo to execute pipeline step pods.

Code Reference

Source Location

Signature

apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: pipeline-runner
rules:
- apiGroups: [""]
  resources: [secrets]
  verbs: [get]
- apiGroups: [""]
  resources: [configmaps]
  verbs: [get, watch, list]
- apiGroups: [""]
  resources: [persistentvolumes, persistentvolumeclaims]
  verbs: ["*"]
- apiGroups: [argoproj.io]
  resources: [workflows, workflowtaskresults]
  verbs: ["*"]
# ... additional rules for pods, deployments, kubeflow resources, etc.

Import

kubectl apply -f pipeline-runner-role.yaml

I/O Contract

Inputs

Name Type Required Description
Kubernetes namespace namespace Yes Target namespace where pipelines execute

Outputs

Name Type Description
RBAC Role Role Grants pipeline step pods access to required K8s resources

Usage Examples

Verifying Pipeline Runner Permissions

# Check the role exists
kubectl get role pipeline-runner

# Verify role binding
kubectl get rolebinding -o wide | grep pipeline-runner

# Test pipeline runner can access PVCs
kubectl auth can-i create persistentvolumeclaims \
  --as=system:serviceaccount:kubeflow:pipeline-runner

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.
Principle
Implementation
Heuristic
Environment