Files
gogo2/COBY/docker/README.md
2025-08-04 15:50:54 +03:00

273 lines
6.3 KiB
Markdown

# Market Data Infrastructure Docker Setup
This directory contains Docker Compose configurations and scripts for deploying TimescaleDB and Redis infrastructure for the multi-exchange data aggregation system.
## 🏗️ Architecture
- **TimescaleDB**: Time-series database optimized for high-frequency market data
- **Redis**: High-performance caching layer for real-time data
- **Network**: Isolated Docker network for secure communication
## 📋 Prerequisites
- Docker Engine 20.10+
- Docker Compose 2.0+
- At least 4GB RAM available for containers
- 50GB+ disk space for data storage
## 🚀 Quick Start
1. **Copy environment file**:
```bash
cp .env.example .env
```
2. **Edit configuration** (update passwords and settings):
```bash
nano .env
```
3. **Deploy infrastructure**:
```bash
chmod +x deploy.sh
./deploy.sh
```
4. **Verify deployment**:
```bash
docker-compose -f timescaledb-compose.yml ps
```
## 📁 File Structure
```
docker/
├── timescaledb-compose.yml # Main Docker Compose configuration
├── init-scripts/ # Database initialization scripts
│ └── 01-init-timescaledb.sql
├── redis.conf # Redis configuration
├── .env # Environment variables
├── deploy.sh # Deployment script
├── backup.sh # Backup script
├── restore.sh # Restore script
└── README.md # This file
```
## ⚙️ Configuration
### Environment Variables
Key variables in `.env`:
```bash
# Database credentials
POSTGRES_PASSWORD=your_secure_password
POSTGRES_USER=market_user
POSTGRES_DB=market_data
# Redis settings
REDIS_PASSWORD=your_redis_password
# Performance tuning
POSTGRES_SHARED_BUFFERS=256MB
POSTGRES_EFFECTIVE_CACHE_SIZE=1GB
REDIS_MAXMEMORY=2gb
```
### TimescaleDB Configuration
The database is pre-configured with:
- Optimized PostgreSQL settings for time-series data
- TimescaleDB extension enabled
- Hypertables for automatic partitioning
- Retention policies (90 days for raw data)
- Continuous aggregates for common queries
- Proper indexes for query performance
### Redis Configuration
Redis is configured for:
- High-frequency data caching
- Memory optimization (2GB limit)
- Persistence with AOF and RDB
- Optimized for order book data structures
## 🔌 Connection Details
After deployment, connect using:
### TimescaleDB
```
Host: 192.168.0.10
Port: 5432
Database: market_data
Username: market_user
Password: (from .env file)
```
### Redis
```
Host: 192.168.0.10
Port: 6379
Password: (from .env file)
```
## 🗄️ Database Schema
The system creates the following tables:
- `order_book_snapshots`: Real-time order book data
- `trade_events`: Individual trade events
- `heatmap_data`: Aggregated price bucket data
- `ohlcv_data`: OHLCV candlestick data
- `exchange_status`: Exchange connection monitoring
- `system_metrics`: System performance metrics
## 💾 Backup & Restore
### Create Backup
```bash
chmod +x backup.sh
./backup.sh
```
Backups are stored in `./backups/` with timestamp.
### Restore from Backup
```bash
chmod +x restore.sh
./restore.sh market_data_backup_YYYYMMDD_HHMMSS.tar.gz
```
### Automated Backups
Set up a cron job for regular backups:
```bash
# Daily backup at 2 AM
0 2 * * * /path/to/docker/backup.sh
```
## 📊 Monitoring
### Health Checks
Check service health:
```bash
# TimescaleDB
docker exec market_data_timescaledb pg_isready -U market_user -d market_data
# Redis
docker exec market_data_redis redis-cli -a your_password ping
```
### View Logs
```bash
# All services
docker-compose -f timescaledb-compose.yml logs -f
# Specific service
docker-compose -f timescaledb-compose.yml logs -f timescaledb
```
### Database Queries
Connect to TimescaleDB:
```bash
docker exec -it market_data_timescaledb psql -U market_user -d market_data
```
Example queries:
```sql
-- Check table sizes
SELECT
schemaname,
tablename,
pg_size_pretty(pg_total_relation_size(schemaname||'.'||tablename)) as size
FROM pg_tables
WHERE schemaname = 'market_data';
-- Recent order book data
SELECT * FROM market_data.order_book_snapshots
ORDER BY timestamp DESC LIMIT 10;
-- Exchange status
SELECT * FROM market_data.exchange_status
ORDER BY timestamp DESC LIMIT 10;
```
## 🔧 Maintenance
### Update Images
```bash
docker-compose -f timescaledb-compose.yml pull
docker-compose -f timescaledb-compose.yml up -d
```
### Clean Up Old Data
```bash
# TimescaleDB has automatic retention policies
# Manual cleanup if needed:
docker exec market_data_timescaledb psql -U market_user -d market_data -c "
SELECT drop_chunks('market_data.order_book_snapshots', INTERVAL '30 days');
"
```
### Scale Resources
Edit `timescaledb-compose.yml` to adjust:
- Memory limits
- CPU limits
- Shared buffers
- Connection limits
## 🚨 Troubleshooting
### Common Issues
1. **Port conflicts**: Change ports in compose file if 5432/6379 are in use
2. **Memory issues**: Reduce shared_buffers and Redis maxmemory
3. **Disk space**: Monitor `/var/lib/docker/volumes/` usage
4. **Connection refused**: Check firewall settings and container status
### Performance Tuning
1. **TimescaleDB**:
- Adjust `shared_buffers` based on available RAM
- Tune `effective_cache_size` to 75% of system RAM
- Monitor query performance with `pg_stat_statements`
2. **Redis**:
- Adjust `maxmemory` based on data volume
- Monitor memory usage with `INFO memory`
- Use appropriate eviction policy
### Recovery Procedures
1. **Container failure**: `docker-compose restart <service>`
2. **Data corruption**: Restore from latest backup
3. **Network issues**: Check Docker network configuration
4. **Performance degradation**: Review logs and system metrics
## 🔐 Security
- Change default passwords in `.env`
- Use strong passwords (20+ characters)
- Restrict network access to trusted IPs
- Regular security updates
- Monitor access logs
- Enable SSL/TLS for production
## 📞 Support
For issues related to:
- TimescaleDB: Check [TimescaleDB docs](https://docs.timescale.com/)
- Redis: Check [Redis docs](https://redis.io/documentation)
- Docker: Check [Docker docs](https://docs.docker.com/)
## 🔄 Updates
This infrastructure supports:
- Rolling updates with zero downtime
- Blue-green deployments
- Automated failover
- Data migration scripts