k8s-launch-kit-validate

Validate NVIDIA Kubernetes networking deployments against source configuration and Network Operator release.

14|5|Updated Nov 4, 2025
One-click install
npx skills add https://github.com/NVIDIA/k8s-launch-kit --skill k8s-launch-kit-validate
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: k8s-launch-kit-validate
Source: https://github.com/NVIDIA/k8s-launch-kit/tree/main/skills/k8s-launch-kit-validate
Command: npx skills add https://github.com/NVIDIA/k8s-launch-kit --skill k8s-launch-kit-validate

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill verifies that an NVIDIA networking deployment is correctly applied to a Kubernetes cluster and matches the configuration and Network Operator release that produced it.

Core Features & Use Cases

  • Release Verification: Compare the deployed Network Operator Helm chart version with the version selected in the cluster configuration.
  • Manifest Validation: Check generated resources for readiness, progress, errors, or missing objects, including operator-populated NIC templates and device selectors.
  • Connectivity Testing: Run source-bound ICMP, RDMA rping, and ib_write_bw checks using quick, full, or strict validation modes.
  • Use Case: After deploying an SR-IOV, RDMA, or Spectrum-X networking profile, use this Skill to identify missing manifests, version drift, topology mismatches, or connectivity failures before troubleshooting further.

Quick Start

Ask the skill to validate the deployment using the cluster configuration and manifests in the default locations, then summarize any version, resource-state, or connectivity failures.

Frequently Asked Questions about k8s-launch-kit-validate

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I validate my Kubernetes Network Operator deployment?

Validate your Kubernetes Network Operator deployment by comparing the deployed Helm chart version against the cluster configuration and checking generated manifests for readiness and missing objects.

Can I test RDMA connectivity in a Kubernetes cluster?

Yes, you can test RDMA connectivity by running source-bound rping and ib_write_bw checks to verify network paths and bandwidth between nodes configured with SR-IOV or Spectrum-X profiles.

How do I verify Spectrum-X networking resources in Kubernetes?

Verify Spectrum-X resources by checking generated manifests for errors, progress, and topology mismatches, ensuring operator-populated NIC templates and device selectors match the expected configuration.

What do I need to check SR-IOV deployment state in Kubernetes?

Checking SR-IOV deployment state requires access to the Kubernetes cluster, deployment manifests, and optionally cluster-config.yaml to perform Helm release validation and source-bound connectivity checks.

Why does my Network Operator Helm release have version drift?

Version drift occurs when the deployed Network Operator Helm chart version does not match the version selected in the source cluster configuration file used to generate the manifests.

What is the best way to troubleshoot missing manifests in an RDMA deployment?

Troubleshoot missing manifests by running a validation check that identifies missing objects, resource readiness issues, and topology mismatches before performing deeper manual diagnostics on the cluster.