orunnerpool 0.1.0

Worker client for the ORunner Pool project

ORunner Pool - Worker Client

[!WARNING] This project is in a very early status. Consider it a _working_✨ proof of concept.

This is the worker component of the ORunner Pool project. It runs on contributor machines to serve Ollama models to the pool.

Prerequisites

• Python 3.10 or higher
• Ollama installed and running on your machine
• At least one model loaded in Ollama
• An account on the Ollama Runner Pool with an API key (https://orunnerpool.com/index.php?page=register)

Installation

1. Clone the repository or download the worker files:

   git clone https://github.com/aaronfc/orunnerpool-worker.git
   cd orunnerpool-worker
   

2. Install the required dependencies:

   pip install -r requirements.txt
   

3. Create your configuration file:

   cp config.ini.example config.ini
   

4. Edit the config.ini file with your settings:

   [auth]
   # Your API key from the Ollama Runner Pool web interface
   api_key = your_api_key_here
   
   [pool]
   # URL of the pool API
   # If using Docker Compose for development, use http://localhost:8000
   # For production, use https://api.orunnerpool.com
   api_url = https://api.orunnerpool.com
   
   [worker]
   # Name of this worker (shown in the web interface)
   name = My Ollama Worker
   # Heartbeat interval in seconds
   heartbeat_interval = 30
   # Task polling interval in seconds
   poll_interval = 5
   
   [ollama]
   # URL of your local Ollama instance
   url = http://localhost:11434
   

Usage

1. Make sure Ollama is running on your machine:

   ollama serve
   

2. Start the worker:

   python worker.py
   

   Or specify a custom config file location:

   python worker.py --config /path/to/your/config.ini
   

3. The worker will automatically:

   • Discover available models on your Ollama instance
   • Register with the pool
   • Send regular heartbeats
   • Poll for tasks and process them

Logs

The worker logs to stdout by default. You can redirect the output to a file:

python worker.py > worker.log 2>&1

Running as a Service

[!WARNING] These steps are not tested yet.

Systemd (Linux)

Create a systemd service file:

sudo nano /etc/systemd/system/ollama-worker.service

Add the following content:

[Unit]
Description=Ollama Runner Pool Worker
After=network.target

[Service]
User=yourusername
WorkingDirectory=/path/to/worker
ExecStart=/usr/bin/python3 /path/to/worker/worker.py
Restart=always
RestartSec=10

[Install]
WantedBy=multi-user.target

Enable and start the service:

sudo systemctl enable ollama-worker
sudo systemctl start ollama-worker

Launchd (macOS)

Create a plist file:

nano ~/Library/LaunchAgents/com.ollama.worker.plist

Add the following content:

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
    <key>Label</key>
    <string>com.ollama.worker</string>
    <key>ProgramArguments</key>
    <array>
        <string>/usr/bin/python3</string>
        <string>/path/to/worker/worker.py</string>
    </array>
    <key>RunAtLoad</key>
    <true/>
    <key>KeepAlive</key>
    <true/>
    <key>WorkingDirectory</key>
    <string>/path/to/worker</string>
    <key>StandardOutPath</key>
    <string>/path/to/worker/worker.log</string>
    <key>StandardErrorPath</key>
    <string>/path/to/worker/worker.log</string>
</dict>
</plist>

Load the service:

launchctl load ~/Library/LaunchAgents/com.ollama.worker.plist

Troubleshooting

Worker fails to register

• Make sure Ollama is running and has at least one model loaded
• Check that your API key is correct
• Verify the pool API URL is accessible

Worker can't connect to Ollama

• Ensure Ollama is running (ollama serve)
• Check the Ollama URL in your config file
• Verify there are no firewall rules blocking access to port 11434

Worker doesn't receive tasks

• Check that your worker is registered successfully
• Verify that your models are correctly listed in the pool
• Make sure your worker's heartbeat is being received by the pool

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.