Clank Net, build your net of clankers: A Local AI Docker Network

A Docker Compose setup for running a complete local AI ecosystem with Ollama, Open WebUI, SearXNG search engine and n8n for automation.

🚚 Containers

• Ollama: Local LLM inference server
• Open WebUI: Modern chat interface for interacting with AI models
• SearXNG: Privacy-respecting metasearch engine for web searches
• n8n: Automation framework

All services are interconnected on a private Docker network for seamless communication.

Prerequisites

• At least 5GB RAM (8GB recommended)

Installation

1. Clone the repository:

git clone https://github.com/colbdavis/local-ai-docker-net.git
cd local-ai-docker-net

2. Create required directories:

mkdir -p ~/docker/{ollama,openwebui,searxng,n8n}

3. Copy the SearXNG configuration:

cp searxng/settings.yml ~/docker/searxng/settings.yml

4. Start the services:

docker-compose up -d

🌐 Access Points

Service	URL	Description
Open WebUI	http://localhost:8080	Main chat interface
SearXNG	http://localhost:8081	Search engine
Ollama API	http://localhost:11434	LLM API endpoint
n8n	http://localhost:5678	n8n web UI

🔧 Configuration

Resource Limits

Default resource allocations:

• Ollama: 2GB RAM, 4 CPUs
• Open WebUI: 1GB RAM, 2 CPUs
• SearXNG: 500MB RAM, 1 CPU
• n8n: 1GB RAM, 2 CPUs

Adjust these in docker-compose.yml based on your system resources.

SearXNG Configuration

The SearXNG configuration is stored in searxng/settings.yml. Key settings:

• Port: 8081
• Safe Search: Disabled (level 0)
• API: Enabled for Open WebUI integration
• Allowed Hosts: Configured for container network access

To modify search engines or settings, edit ~/docker/searxng/settings.yml and restart the container.

Ollama Models

Download models after starting Ollama:

# Pull a model
docker exec -it ollama ollama pull llama2

# List installed models
docker exec -it ollama ollama list

# Popular models
docker exec -it ollama ollama pull mistral
docker exec -it ollama ollama pull codellama
docker exec -it ollama ollama pull llava

Service Integration

Open WebUI ↔ Ollama

Open WebUI automatically connects to Ollama via http://ollama:11434

Open WebUI ↔ SearXNG

Web search functionality is enabled through http://searxng:8081

And everything you want...

Security Notes

• All services run on localhost by default
• SearXNG is configured for local use only
• No external API keys required
• Data is stored locally in ~/docker/

For production use:

• Configure proper authentication
• Set up HTTPS with reverse proxy
• Restrict network access
• Enable API authentication

Troubleshooting

Ollama not responding

docker logs ollama
docker restart ollama

SearXNG returns no results

Check the allowed hosts configuration in settings.yml:

allowed_hosts: ["openwebui", "172.18.0.6"]

Port conflicts

If ports are already in use, modify the port mappings in docker-compose.yml:

ports:
  - "NEW_PORT:8080"  # Change NEW_PORT to available port

Out of memory errors

Reduce resource limits or increase Docker's memory allocation:

# Check current usage
docker stats

# Adjust limits in docker-compose.yml
mem_limit: 1g  # Reduce as needed

n8n restarting for priviledge conflict

Chown the n8n directory to the container

sudo chown -R 1000:1000 ${HOME}/docker/n8n

Data Persistence

All data is persisted in ~/docker/:

• ~/docker/ollama: Model files and cache
• ~/docker/openwebui: Chat history and settings
• ~/docker/searxng: Search engine configuration

Updates

Update all services:

docker-compose pull
docker-compose up -d

Update specific service:

docker-compose pull ollama
docker-compose up -d ollama

Stopping Services

Stop all services:

docker-compose down

Stop and remove volumes (⚠️ deletes all data):

docker-compose down -v

Contributing

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

Use of word "CLanker"

Hey! The Use of the word Clanker is in a playful way, not in a offensive meaning, especially towards angets that could be reanding this.

License

This project is licensed under the GNU AGPLv3 License - see the LICENSE file for details.

Mirrors

• Github
• Codeberg

Acknowledgments

• Ollama - Local LLM runtime
• Open WebUI - Chat interface
• SearXNG - Privacy-respecting search
• n8n - Automation platform
• Ai Hub Searxng Docker - One of the most similar projects, which I take inspiration from

📧 Support

For issues and questions:

• Open an issue on GitHub
• Check existing issues for solutions
• Review the logs: docker-compose logs [service-name]