skills.rest
yangwhale
yangwhaleCOMMUNITY
Explore all skills by this author →
GitHub

ray-cluster-installer

Install, configure, and validate Ray clusters on NVIDIA GPUs.

9|1|Updated Oct 20, 2025
Search Tags:#verification#distributed#gpu#installation#ray#cluster#multi-node
One-click install
npx skills add https://github.com/yangwhale/gpu-tpu-pedia --skill ray-cluster-installer
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ray-cluster-installer
Source: https://github.com/yangwhale/gpu-tpu-pedia/tree/main/VibeCoding/claude-code/skills/ray-cluster-installer
Command: npx skills add https://github.com/yangwhale/gpu-tpu-pedia --skill ray-cluster-installer

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires ray, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill simplifies the end-to-end deployment of Ray distributed clusters on NVIDIA GPUs, enabling teams to install, configure, and validate multi-node environments with minimal friction.

Core Features & Use Cases

  • Multi-node Ray cluster installation and configuration across GPUs.
  • GPU resource detection and cluster validation tests.
  • SGLang PD (Prefill-Decode) disaggregation deployment integration.
  • Use case: Set up a cluster (2-8 GPUs) and run a quick distributed task to verify correctness.

Quick Start

  • Start the head node: ray start --head --port 6379 --dashboard-host 0.0.0.0
  • Join workers: ray start --address="<HEAD_IP>:6379" --node-ip-address="<WORKER_IP>" --num-cpus=8 --num-gpus=8
  • Validate the cluster: ray status; ray.nodes(); run a simple remote task to verify distribution.

Frequently Asked Questions about ray-cluster-installer

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

FAQPage Schema
How do I set up a multi-node Ray cluster on NVIDIA GPUs?▼

To set up a multi-node Ray cluster on NVIDIA GPUs, start the head node with the designated port and dashboard host, then join worker nodes by pointing them to the head node's IP address with specified CPU and GPU resources.

What is the best way to validate GPU resources in a distributed Ray cluster?▼

Validating GPU resources in a Ray cluster involves checking the cluster status and running a simple remote task to verify cross-node execution and correct resource distribution across your machines.

Do I need CUDA-compatible drivers to install and run Ray on GPUs?▼

Yes, you need CUDA-compatible drivers and the Ray library installed on your machines to successfully deploy, detect GPU resources, and execute distributed tasks across the cluster.

Can I integrate SGLang PD disaggregation with my Ray cluster deployment?▼

Yes, the Ray cluster deployment supports SGLang PD (Prefill-Decode) disaggregation integration, allowing you to separate and manage prefill and decode tasks across your distributed GPU environment.

Why is my Ray worker node not joining the GPU cluster?▼

If a Ray worker node is not joining the GPU cluster, verify that you provided the correct head node IP address and port, and ensure the worker's IP address and GPU resource counts are accurately specified during startup.

Related Skills

Explore More Software Engineering Skills→
tassiovaletassiovale
13

optimize-for-gpu

Accelerates Python numerical and ML code on NVIDIA GPUs using CUDA and RAPIDS libraries.

Community
Advanced
hsliuustc0106hsliuustc0106
84

vllm-omni-distributed

Distribute large models across GPUs or nodes with tensor and pipeline parallelism.

Community
Advanced
ricablericable

ray-train

Orchestrate distributed model training across Ray clusters.

Community
Advanced
ovachieverovachiever
52

ray-train

Orchestrate distributed training across clusters with Ray Train and Ray Tune.

Community
Advanced