rhoai-kubeflow-spark-operator

name	rhoai-kubeflow-spark-operator
metadata	{"author":"rhoai3-demo","version":"1.0.0","platform-family":"rhoai","platform-baseline":"repo","ocp-baseline":"repo","skill-group":"RHOAI Platform"}
description	Use when documenting, reviewing, or rebuilding Red Hat OpenShift AI Kubeflow Spark Operator workflows from the official guide: KSO architecture, activating kubeflowsparkoperator in the DataScienceCluster, verifying the operator pod in redhat-ods-applications, building or importing an OpenShift-compatible Spark 4.0.1+ image, authoring SparkApplication resources, handling Spark driver and executor resources, removing runAsGroup/runAsUser from example YAML, copying Spark Operator RBAC into a custom namespace, and classifying sparkoperator.k8s.io/v1beta2 API stability. Do NOT use for general distributed training with Ray/Kueue/PyTorchJob (use rhoai-distributed-workload-workflows), AI Pipelines/KFP workflows (use rhoai-ai-pipelines and rhoai-kfp-pipeline-authoring), generic project lifecycle (use rhoai-project-workflows), generic image build policy (use project-gitops-authoring and project-manifest-review), or live cluster changes without the OpenShift safety guard.

RHOAI Kubeflow Spark Operator

Use this skill for Red Hat OpenShift AI Kubeflow Spark Operator workflows on the active product baseline in docs/PLATFORM_BASELINE.md.

Source Grounding

Read references/source-capture.md before using product workflow details. Official Red Hat documentation is product authority. This skill adapts the official Kubeflow Spark Operator guide to this repo's GitOps and demo workflow model.

Scope

This skill covers:

• Kubeflow Spark Operator purpose and architecture
• activation through the OpenShift AI DataScienceCluster custom resource
• verification of the Spark Operator pod in