Docker Setup Guide for SmartRAG Demo

Complete Docker setup for SmartRAG Demo with all local services

Complete guide for running SmartRAG with all local services using Docker.

🐳 Docker Compose Configuration

Service Details

Service	Image	Container	Ports	Volume	Health Check
SQL Server	`mcr.microsoft.com/mssql/server:2022-latest`	`smartrag-sqlserver-test`	`1433:1433`	`sqlserver-data`	SQL query test
MySQL	`mysql:8.0`	`smartrag-mysql-test`	`3306:3306`	`mysql-data`	MySQL ping test
PostgreSQL	`postgres:16-alpine`	`smartrag-postgres-test`	`5432:5432`	`postgres-data`	PostgreSQL ready test
Ollama	`ollama/ollama:latest`	`smartrag-ollama`	`11434:11434`	`ollama-data`	Ollama list test
Qdrant	`qdrant/qdrant:latest`	`smartrag-qdrant`	`6333:6333`, `6334:6334`	`qdrant-data`	HTTP health check
Redis	`redis/redis-stack-server:latest`	`smartrag-redis`	`6379:6379`, `8001:8001`	`redis-data`	Redis ping test

Volume Persistence

All data persists across container restarts using Docker volumes:

sqlserver-data - SQL Server database files
mysql-data - MySQL database files
postgres-data - PostgreSQL database files
ollama-data - Ollama AI models and cache
qdrant-data - Qdrant vector database files
redis-data - Redis cache and persistence files

Health Checks

All services include health checks with automatic restart on failure:

Interval: 10-30 seconds (depending on service)
Timeout: 5-10 seconds
Retries: 3-5 attempts
Restart Policy: unless-stopped

📦 What Gets Installed

This Docker setup includes 6 services for complete SmartRAG functionality:

AI & Storage

Ollama - Local AI models (LLaMA, Phi, Mistral, etc.)
Qdrant - Vector database for embeddings
Redis - Cache and vector storage

Databases

SQL Server 2022 Express - Primary database for demo data
MySQL 8.0 - Secondary database for demo data
PostgreSQL 16 - Tertiary database for demo data

🚀 Quick Start

Why Docker? Get SmartRAG running in minutes with zero configuration. All databases, AI models, and vector stores are pre-configured and ready to use.

1. Environment Setup (Recommended)

For better security, use environment variables instead of default passwords:

# Copy the example file
cp env.example .env

# Edit .env with your secure passwords
nano .env  # or use your preferred editor

Example .env file:

# SQL Server Configuration
SQLSERVER_SA_PASSWORD=YourSecurePassword123!

# MySQL Configuration  
MYSQL_ROOT_PASSWORD=YourMySQLPassword456!

# PostgreSQL Configuration
POSTGRES_PASSWORD=YourPostgresPassword789!

2. Start All Services

cd examples/SmartRAG.Demo
docker-compose up -d

Wait 10-15 seconds for all services to initialize.

3. Verify Services

docker-compose ps

You should see all 6 containers running with "healthy" status.

4. Setup Ollama Models

# Download LLaMA 3.2 (main AI model)
docker exec -it smartrag-ollama ollama pull llama3.2

# Download embedding model (required for RAG)
docker exec -it smartrag-ollama ollama pull nomic-embed-text

Note: First download may take 5-10 minutes depending on your internet speed.

5. Run the Application

dotnet run

Select "1. LOCAL Environment" when prompted.

🔧 Individual Service Management

Start Individual Services

# AI Services
docker-compose up -d smartrag-ollama
docker-compose up -d smartrag-qdrant
docker-compose up -d smartrag-redis

# Databases
docker-compose up -d smartrag-sqlserver
docker-compose up -d smartrag-mysql
docker-compose up -d smartrag-postgres

Stop Services

# Stop all
docker-compose stop

# Stop specific service
docker-compose stop smartrag-ollama

Restart Services

# Restart all
docker-compose restart

# Restart specific
docker-compose restart smartrag-qdrant

View Logs

# All services
docker-compose logs

# Specific service
docker-compose logs smartrag-ollama
docker-compose logs smartrag-qdrant
docker-compose logs smartrag-redis

# Follow logs (real-time)
docker-compose logs -f smartrag-ollama

📊 Connection Details

Ollama (AI)

Endpoint: http://localhost:11434
API: http://localhost:11434/api/tags
Container: smartrag-ollama
Volume: ollama-data

Test Connection:

curl http://localhost:11434/api/tags

Qdrant (Vector Database)

HTTP: http://localhost:6333
gRPC: localhost:6334
Dashboard: http://localhost:6333/dashboard
Container: smartrag-qdrant
Volume: qdrant-data

Note: SmartRAG Dashboard (document management + chat) is at your app URL + /smartrag, not Qdrant's dashboard.

Test Connection:

curl http://localhost:6333/health

Redis (Cache & Vector Store)

Host: localhost:6379
RedisInsight UI: http://localhost:8001 (optional web UI)
Container: smartrag-redis
Volume: redis-data
Persistence: AOF (Append Only File) enabled
RediSearch: Included (vector similarity search enabled)

Test Connection:

docker exec -it smartrag-redis redis-cli ping
# Should return: PONG

# Verify RediSearch module
docker exec -it smartrag-redis redis-cli MODULE LIST
# Should show: search

SQL Server

Host: localhost,1433
Username: sa
Password: ${SQLSERVER_SA_PASSWORD} (environment variable required)
Database: PrimaryDemoDB (created by app)
Container: smartrag-sqlserver-test
Volume: sqlserver-data

Test Connection:

# Using environment variable
export SQLSERVER_SA_PASSWORD="your_secure_password"
docker exec -it smartrag-sqlserver-test /opt/mssql-tools18/bin/sqlcmd -S localhost -U sa -P $SQLSERVER_SA_PASSWORD -C -Q "SELECT @@VERSION"

MySQL

Host: localhost:3306
Username: root
Password: ${MYSQL_ROOT_PASSWORD} (environment variable required)
Database: SecondaryDemoDB (created by app)
Container: smartrag-mysql-test
Volume: mysql-data

Test Connection:

docker exec -it smartrag-mysql-test mysql -u root -p${MYSQL_ROOT_PASSWORD} -e "SELECT VERSION()"

PostgreSQL

Host: localhost:5432
Username: postgres
Password: ${POSTGRES_PASSWORD} (environment variable required)
Database: TertiaryDemoDB (created by app)
Container: smartrag-postgres-test
Volume: postgres-data

Test Connection:

docker exec -it smartrag-postgres-test psql -U postgres -c "SELECT version()"

🤖 Ollama Model Management

List Installed Models

docker exec -it smartrag-ollama ollama list

Pull/Download Models

# Recommended models
docker exec -it smartrag-ollama ollama pull llama3.2        # Main AI (3.2GB)
docker exec -it smartrag-ollama ollama pull llama3.2:1b     # Lightweight (1.3GB)
docker exec -it smartrag-ollama ollama pull nomic-embed-text # Embeddings (274MB)
docker exec -it smartrag-ollama ollama pull phi3            # Microsoft Phi (2.3GB)
docker exec -it smartrag-ollama ollama pull mistral         # Mistral 7B (4.1GB)
docker exec -it smartrag-ollama ollama pull qwen2.5         # Alibaba Qwen (4.7GB)

Remove Models

docker exec -it smartrag-ollama ollama rm llama3.2

Check Model Info

docker exec -it smartrag-ollama ollama show llama3.2

💾 Data Persistence

All data is stored in Docker volumes and persists across container restarts.

List Volumes

docker volume ls | grep smartrag

Inspect Volume

docker volume inspect smartrag-localdemo_ollama-data
docker volume inspect smartrag-localdemo_qdrant-data
docker volume inspect smartrag-localdemo_redis-data

Backup Data

# Backup Ollama models
docker run --rm -v smartrag-localdemo_ollama-data:/data -v $(pwd):/backup alpine tar czf /backup/ollama-backup.tar.gz -C /data .

# Backup Qdrant
docker run --rm -v smartrag-localdemo_qdrant-data:/data -v $(pwd):/backup alpine tar czf /backup/qdrant-backup.tar.gz -C /data .

Restore Data

# Restore Ollama
docker run --rm -v smartrag-localdemo_ollama-data:/data -v $(pwd):/backup alpine sh -c "cd /data && tar xzf /backup/ollama-backup.tar.gz"

🧹 Cleanup

Remove Containers (Keep Data)

docker-compose down

Remove Everything (⚠️ Including Data)

docker-compose down -v

Remove Specific Volume

docker volume rm smartrag-localdemo_ollama-data

Clean Up Unused Docker Resources

docker system prune -a

🔧 Troubleshooting

Port Already in Use

If a port is occupied, change it in docker-compose.yml:

# Example: Change Ollama port from 11434 to 11435
ollama:
  ports:
    - "11435:11434"  # Host:Container

Then update appsettings.Development.json:

{
  "AI": {
    "Custom": {
      "Endpoint": "http://localhost:11435"
    }
  }
}

Container Won't Start

# Check Docker is running
docker --version

# Check container logs
docker logs smartrag-ollama

# Restart Docker Desktop
# (Sometimes needed on Windows)

Out of Disk Space

# Check Docker disk usage
docker system df

# Remove unused images
docker image prune -a

# Remove unused volumes
docker volume prune

Health Check Failing

# Check health status
docker inspect smartrag-ollama | grep -A 10 Health

# Manually test health endpoint
curl http://localhost:11434/api/tags   # Ollama
curl http://localhost:6333/health      # Qdrant
docker exec -it smartrag-redis redis-cli ping  # Redis

Ollama Model Download Stuck

# Check Ollama logs
docker logs -f smartrag-ollama

# Restart Ollama
docker-compose restart smartrag-ollama

# Try pulling again
docker exec -it smartrag-ollama ollama pull llama3.2

📊 Resource Usage

Typical Resource Consumption

Service	RAM	Disk
Ollama (idle)	~200MB	~500MB base
Ollama + llama3.2	~2-4GB	~3.2GB
Ollama + multiple models	~4-8GB	~10-15GB
Qdrant	~100MB	~50MB + data
Redis	~50MB	~10MB + cache
SQL Server	~2GB	~500MB
MySQL	~300MB	~200MB
PostgreSQL	~50MB	~100MB

Recommended System:

RAM: 8GB minimum, 16GB recommended
Disk: 20GB free space
CPU: 4+ cores recommended for AI inference

Monitor Resource Usage

# Real-time stats
docker stats

# Check specific container
docker stats smartrag-ollama

🔒 Security Notes

⚠️ This setup is for local development/testing only!

Default passwords are intentionally simple
Services are exposed on localhost only
NEVER use these configurations in production
NEVER expose these ports to the internet

Environment Variables for Security

For better security, use environment variables instead of hardcoded passwords:

# Set secure passwords
export SQLSERVER_SA_PASSWORD="YourSecurePassword123!"
export MYSQL_ROOT_PASSWORD="YourMySQLPassword456!"
export POSTGRES_PASSWORD="YourPostgresPassword789!"

# Start with custom passwords
docker-compose up -d

For production deployments:

Use strong, unique passwords via environment variables
Enable TLS/SSL encryption
Use proper authentication
Run behind a firewall
Use Docker secrets for sensitive data

🌐 Alternative: Manual Installation

If you prefer not to use Docker:

Ollama

Download from: https://ollama.ai
Install locally
Models stored in: ~/.ollama/models

Qdrant

Download from: https://qdrant.tech/documentation/guides/installation/
Or use Qdrant Cloud: https://cloud.qdrant.io

Redis

Important: For vector search, use redis/redis-stack-server (includes RediSearch module)
Docker: docker run -d -p 6379:6379 redis/redis-stack-server:latest
Windows: https://github.com/microsoftarchive/redis/releases (requires RediSearch module installation)
Linux: sudo apt install redis-server (requires RediSearch module installation)
macOS: brew install redis (requires RediSearch module installation)
Note: Standard Redis does not include RediSearch. Without RediSearch, only text search will work (no vector similarity search).

📚 Additional Resources

Ollama Documentation: https://ollama.ai/docs
Qdrant Documentation: https://qdrant.tech/documentation/
Redis Documentation: https://redis.io/documentation
SmartRAG Documentation: https://byerlikaya.github.io/SmartRAG/en/

🤝 Contact

For issues or questions:

GitHub: https://github.com/byerlikaya/SmartRAG
LinkedIn: https://www.linkedin.com/in/barisyerlikaya/
NuGet: https://www.nuget.org/packages/SmartRAG
Website: https://byerlikaya.github.io/SmartRAG/en/
Email: b.yerlikaya@outlook.com

FilesExpand file tree

README-Docker.md

Latest commit

History

README-Docker.md

File metadata and controls

Docker Setup Guide for SmartRAG Demo

🐳 Docker Compose Configuration

Service Details

Volume Persistence

Health Checks

📦 What Gets Installed

AI & Storage

Databases

🚀 Quick Start

1. Environment Setup (Recommended)

2. Start All Services

3. Verify Services

4. Setup Ollama Models

5. Run the Application

🔧 Individual Service Management

Start Individual Services

Stop Services

Restart Services

View Logs

📊 Connection Details

Ollama (AI)

Qdrant (Vector Database)

Redis (Cache & Vector Store)

SQL Server

MySQL

PostgreSQL

🤖 Ollama Model Management

List Installed Models

Pull/Download Models

Remove Models

Check Model Info

💾 Data Persistence

List Volumes

Inspect Volume

Backup Data

Restore Data

🧹 Cleanup

Remove Containers (Keep Data)

Remove Everything (⚠️ Including Data)

Remove Specific Volume

Clean Up Unused Docker Resources

🔧 Troubleshooting

Port Already in Use

Container Won't Start

Out of Disk Space

Health Check Failing

Ollama Model Download Stuck

📊 Resource Usage

Typical Resource Consumption

Monitor Resource Usage

🔒 Security Notes

Environment Variables for Security

🌐 Alternative: Manual Installation

Ollama

Qdrant

Redis

📚 Additional Resources

🤝 Contact