autoSMART v1.0 - Intelligent HDD Monitoring & Failure Prediction

autoSMART este un sistem inteligent de monitorizare SMART pentru HDD-urile din cluster-ul Proxmox, cu predicții de defectare bazate pe AI și stocare optimizată în PostgreSQL.

🎯 Scopul Proiectului

Monitorizare continuă a parametrilor SMART pentru toate HDD-urile din cluster
Predicții AI pentru defectări iminente folosind OpenAI API
Stocare long-term în PostgreSQL pentru analize temporale
Alerting proactiv pentru mentenanță preventivă

Key Features

🔍 Hardware-based HDD tracking: Permanent identification using serial numbers and model names (not volatile /dev/sdX paths)
🔄 Migration detection: Automatic detection and logging when HDDs move between nodes or device paths
💾 Differential storage optimization: Store only SMART readings with changes, reducing database size by 60-80%
🤖 AI-powered failure prediction: Uses OpenAI GPT for intelligent drive failure forecasting
🏥 Health monitoring: Continuous SMART parameter analysis with configurable thresholds
📊 Comprehensive reporting: Detailed drive health reports and predictive analytics
🔧 Proxmox cluster integration: Designed for distributed Proxmox VE environments
⚡ High performance: PostgreSQL backend with optimized indexing and queries

🚀 Quick Start

Prerequisites

PostgreSQL 13+ for data storage
Perl 5.20+ with required modules
Proxmox VE cluster environment
smartmontools for SMART data collection
OpenAI API key for failure predictions

Installation

# 1. Download autoSMART and run automated deployment
git clone <repository-url>
cd autoSMART
sudo ./scripts/deploy.sh install

# The deployment script automatically:
# - Installs all dependencies (Perl modules, smartmontools, etc.)
# - Creates system directories and sets permissions
# - Deploys application files to /opt/autoSMART/
# - Creates configuration files in /etc/autosmart/
# - Registers and starts systemd services
# - Performs initial system validation

# 2. Configure database connection (interactive prompts during install)
# 3. Configure OpenAI API key (interactive prompts during install)
# 4. System is ready - services are automatically started

Verification

# Check system status (all services should be active)
sudo systemctl status autosmart

# View recent SMART data collection
sudo journalctl -u autosmart-collector -f

# Generate initial health report
sudo /opt/autoSMART/scripts/autosmart-report.pl --summary

📚 Documentation

Getting Started

CHANGELOG.md - Version history and release notes

System Configuration

API.md - OpenAI API integration and configuration

🏥 Monitoring Dashboard

autoSMART provides comprehensive monitoring capabilities:

Health Status Overview

Real-time drive health status for all cluster nodes
Critical parameter alerts and warnings
AI-powered failure predictions with confidence scores
Storage efficiency metrics

Historical Analysis

Long-term SMART parameter trends
Performance degradation tracking
Migration history between nodes
Predictive analytics reports

Alerting System

Configurable thresholds for all SMART parameters
Email/webhook notifications
Integration with monitoring systems
Escalation procedures for critical alerts

🔧 System Architecture

autoSMART operates as a distributed system across your Proxmox cluster:

Data Collection

Continuous SMART data collection from all nodes
Hardware-based drive identification
Migration detection and logging
Differential storage for efficiency

Analysis Engine

AI-powered failure prediction
Threshold-based alerting
Trend analysis and reporting
Performance optimization recommendations

Storage Layer

PostgreSQL database with optimized schema
Differential storage reducing size by 60-80%
Historical data retention policies
Automated backup and maintenance

📁 Installed File Structure

When autoSMART is installed on your system, it creates the following directory structure:

System Directories

/opt/autoSMART/                    # Main installation directory
├── scripts/                      # Executable scripts and utilities
│   ├── autosmart-collector.pl    # Main data collection daemon
│   ├── autosmart-predictor.pl    # AI prediction processing
│   ├── autosmart-report.pl       # Report generation engine
│   ├── autosmart-migration-report.pl # Hardware migration analysis
│   ├── smart-collector-daemon.pl # Background collection service
│   ├── uninstall.sh             # System removal script
│   ├── monitor-cluster.sh        # Cluster health monitoring
│   └── test-*.pl                # Testing and validation utilities
├── lib/                         # Perl modules and core libraries
│   ├── SmartCollector.pm        # SMART data collection and hardware tracking
│   └── PredictionEngine.pm      # AI-powered failure prediction engine
├── config/                      # Configuration templates and examples
│   └── (template files)        # Default configuration templates
├── docs/                        # End-user documentation
│   ├── README.md               # System overview and quick start
│   ├── CHANGELOG.md            # Release notes and version history
│   └── API.md                  # OpenAI API configuration guide

/etc/autosmart/                   # System configuration directory
├── autosmart.conf              # Main system configuration
├── cluster.conf                # Cluster topology and node definitions
├── database.conf               # PostgreSQL connection settings
├── openai.conf                 # OpenAI API configuration and prompts
└── smart.conf                  # SMART parameter thresholds and monitoring rules

/etc/systemd/system/             # Systemd service files
├── autosmart.service           # Main autoSMART service
├── autosmart-collector.service # Data collection service
└── autosmart-predictor.service # AI prediction service

Configuration Files Detail

`/etc/autosmart/autosmart.conf`

Main system configuration file containing: - Database connection parameters - Collection intervals and scheduling - Local node identification and settings - Log levels and debugging options

`/etc/autosmart/cluster.conf`

Cluster-wide configuration shared across all nodes: - Node topology and IP addresses - Shared monitoring parameters - Cluster-wide alert settings - Inter-node communication settings

`/etc/autosmart/database.conf`

PostgreSQL database connection settings: - Database host, port, and credentials - Connection pooling configuration - SSL settings and security parameters - Performance tuning options

`/etc/autosmart/openai.conf`

OpenAI API integration configuration: - API key and model selection - Prompt templates for failure prediction - Response parsing and confidence thresholds - Rate limiting and cost management

`/etc/autosmart/smart.conf`

SMART parameter monitoring configuration: - Parameter thresholds for different drive types - Critical parameter definitions - Alert escalation rules and notifications - Drive-specific monitoring settings

Service Integration

Systemd Services

autosmart.service: Main system service that manages other components
autosmart-collector.service: Background data collection service
autosmart-predictor.service: AI prediction processing service

Service Management

# Start/stop services
sudo systemctl start autosmart
sudo systemctl stop autosmart

# Enable/disable automatic startup
sudo systemctl enable autosmart
sudo systemctl disable autosmart

# Check service status
sudo systemctl status autosmart

# View service logs using systemd journal
sudo journalctl -u autosmart -f                    # Follow main service logs
sudo journalctl -u autosmart-collector -f          # Follow data collection logs  
sudo journalctl -u autosmart-predictor -f          # Follow AI prediction logs

# View logs by time period
sudo journalctl -u autosmart --since "1 hour ago"  # Last hour
sudo journalctl -u autosmart --since today         # Today's logs
sudo journalctl -u autosmart --since yesterday     # Yesterday's logs

# View logs by priority level
sudo journalctl -u autosmart -p err                # Error level and above
sudo journalctl -u autosmart -p warning            # Warning level and above

File Permissions

Executable Files

All scripts in /opt/autoSMART/scripts/ are executable (755)
Perl modules in /opt/autoSMART/lib/ are readable (644)
Configuration files in /etc/autosmart/ are readable by autosmart user (640)

Log Management

All application logs are handled by systemd journal
No separate log files created in filesystem
Log retention managed by journald configuration
Logs accessible via journalctl commands
Automatic log rotation and cleanup by systemd

Storage Requirements

Disk Space

Installation: ~50MB for application files and documentation
Configuration: ~1MB for all configuration files
Logs: Managed by systemd journal (configurable retention)
Database: Handled separately on PostgreSQL server

Network Requirements

Database Access: Persistent connection to PostgreSQL server
OpenAI API: HTTPS access for AI predictions (configurable)
Inter-node Communication: SSH access between cluster nodes for deployment

This file structure provides a complete, organized installation that integrates seamlessly with Linux system conventions while maintaining clear separation between application code, configuration, and operational data.

📊 Performance Benefits

Storage Optimization

60-80% reduction in database storage through differential storage
Intelligent change detection stores only modified SMART parameters
Baseline reconstruction provides complete historical views
Configurable retention policies for long-term storage

Monitoring Efficiency

Hardware-based tracking eliminates /dev/sdX path volatility
Migration detection automatically tracks drive movements
Real-time analysis with configurable collection intervals
Distributed architecture scales across cluster nodes

🚨 Alert Examples

Critical Alerts

Imminent Failure: AI predicts drive failure within 24-48 hours
Temperature Critical: Drive operating above safe temperature thresholds
Reallocated Sectors: Increasing bad sector count detected
Spin Retry Count: Mechanical issues detected

Warning Alerts

Performance Degradation: Slower response times detected
Temperature Warning: Operating temperatures approaching limits
SMART Threshold: Parameters approaching warning thresholds
Migration Detected: Drive moved to different node or path

💡 Use Cases

Preventive Maintenance

Schedule drive replacements before failures occur
Optimize workload distribution based on drive health
Plan cluster maintenance windows effectively
Track warranty and replacement schedules

Capacity Planning

Monitor storage growth trends
Predict future storage requirements
Optimize drive allocation across nodes
Plan cluster expansion timing

Performance Optimization

Identify performance bottlenecks
Balance load across healthy drives
Optimize I/O patterns based on drive characteristics
Monitor storage tier performance

🆘 Support & Troubleshooting

Common Issues

Collection failures: Check smartmontools installation
Database connectivity: Verify PostgreSQL connection settings
API errors: Validate OpenAI API key and quotas
Performance issues: Review differential storage configuration

Log Analysis

Use systemd journal for comprehensive log analysis: - All service logs: sudo journalctl -u autosmart* - Data collection: sudo journalctl -u autosmart-collector - AI predictions: sudo journalctl -u autosmart-predictor - System errors: sudo journalctl -u autosmart* -p err

Getting Help

For detailed installation, configuration, and troubleshooting information, refer to the complete documentation in the docs/ directory.

autoSMART v1.0 - Intelligent drive monitoring for mission-critical infrastructure