Skip to content

Milaskinger/ghostnexus-node

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ghostnexus-node — Provider Node

PyPI version Python 3.9+ License: MIT

Share your GPU on the GhostNexus decentralized cloud and earn passive income.

pip install ghostnexus-node
export GHOSTNEXUS_PROVIDER_API_KEY="gn_live_YOUR_KEY"
ghostnexus-node

RTX 4090 earns ~$0.35/hr at $0.50/hr pricing (70% payout). 24/7 passive income from idle GPU time.


💰 Earnings Estimate

GPU GhostNexus rate Your cut (70%) Monthly (24/7)
RTX 4090 $0.50/hr $0.35/hr ~$252
RTX 4080 $0.40/hr $0.28/hr ~$202
RTX 3090 $0.30/hr $0.21/hr ~$151
A100 40GB $1.50/hr $1.05/hr ~$756

Earnings depend on utilization. Sign up at ghostnexus.net/start.


🚀 Quick Start

1. Install

pip install ghostnexus-node

2. Get your provider API key

Sign up at ghostnexus.net and go to Dashboard → Provider → Get API Key.

3. Run

export GHOSTNEXUS_PROVIDER_API_KEY="gn_live_YOUR_KEY"
ghostnexus-node

⚙️ Configuration

Variable Default Description
GHOSTNEXUS_PROVIDER_API_KEY (required) Your provider API key
GHOSTNEXUS_WS_URL wss://ghostnexus.net/ws/provider WebSocket endpoint
GHOSTNEXUS_DOCKER_IMAGE python:3.11-slim Docker image for job sandboxing
GHOSTNEXUS_JOB_MEMORY 512m Memory limit per job
GHOSTNEXUS_JOB_CPUS 1.0 CPU limit per job
GHOSTNEXUS_GPU_PASSTHROUGH false Pass GPU into containers
GHOSTNEXUS_JOB_TIMEOUT 300 Max job duration (seconds)
LOG_LEVEL INFO Logging verbosity

Recommended for RTX GPU providers

export GHOSTNEXUS_GPU_PASSTHROUGH=true
export GHOSTNEXUS_JOB_MEMORY=8g
export GHOSTNEXUS_JOB_CPUS=4.0
export GHOSTNEXUS_JOB_TIMEOUT=3600

🛡️ Security

Jobs run in Docker containers with strict sandboxing:

  • --network=none — no internet access
  • --read-only — immutable filesystem
  • --cap-drop=ALL — no Linux capabilities
  • --security-opt no-new-privileges:true — no privilege escalation
  • --user 65534 — runs as nobody
  • Hard memory + CPU limits via cgroups

Prerequisites: Python 3.9+, Docker, NVIDIA GPU + drivers, NVIDIA Container Toolkit (for GPU passthrough)


🐧 Linux systemd service

[Unit]
Description=GhostNexus Provider Node
After=network.target docker.service

[Service]
Type=simple
Environment=GHOSTNEXUS_PROVIDER_API_KEY=gn_live_YOUR_KEY
Environment=GHOSTNEXUS_GPU_PASSTHROUGH=true
Environment=GHOSTNEXUS_JOB_MEMORY=8g
Environment=GHOSTNEXUS_JOB_CPUS=4.0
ExecStart=/usr/local/bin/ghostnexus-node
Restart=always
RestartSec=10

[Install]
WantedBy=multi-user.target
sudo systemctl enable ghostnexus-node && sudo systemctl start ghostnexus-node

🪟 Windows auto-start

Create start_provider.bat:

@echo off
set GHOSTNEXUS_PROVIDER_API_KEY=gn_live_YOUR_KEY
set GHOSTNEXUS_GPU_PASSTHROUGH=true
set GHOSTNEXUS_JOB_MEMORY=8g
pythonw -m ghostnexus_node

Add to shell:startup for auto-run on boot.


🤝 Contributing

git clone https://github.com/Milaskinger/ghostnexus-node
cd ghostnexus-node
pip install -e ".[dev]"

🔗 Links


📄 License

MIT — see LICENSE.

About

GhostNexus provider node — share your GPU and earn on the decentralized GPU cloud

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages