Complete rewrite: SnapRAID SMART logger with BackBlaze algorithm

Fetches SMART data from Scrutiny's InfluxDB and calculates failure probabilities using the exact BackBlaze tables from SnapRAID source. Key features: - Calculates both v12 (with attr 193) and v13 (without attr 193) algorithms - v12 matches SnapRAID pre-v13.0 (includes Load Cycle Count) - v13 matches modern SnapRAID v13.0+ (excludes Load Cycle Count) - Stores both values for trending and comparison - Correctly applies bit masks (16-bit for 187/188, 32-bit for others) - Annualizes monthly rates and applies Poisson distribution - Logs to PostgreSQL with device metadata Solved the mystery: SnapRAID v12 uses attribute 193 which was removed in v13.0. Load cycle count has massive impact on failure predictions for some drives (80% vs 4% difference). 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
2025-12-07 06:55:39 -05:00
commit 27afedb32e
25 changed files with 3207 additions and 0 deletions
@@ -0,0 +1,182 @@
+# SnapRAID SMART Logger
+
+A Python utility that collects SMART data from Scrutiny's InfluxDB, calculates BackBlaze-based failure probabilities (like SnapRAID does), and logs the enhanced data to PostgreSQL.
+
+## Features
+
+- ✅ Connects to Scrutiny's InfluxDB to fetch SMART attributes
+- ✅ Implements SnapRAID's BackBlaze failure probability algorithm
+- ✅ Calculates annual failure rate (AFR) for each disk
+- ✅ Stores historical data in PostgreSQL with time-series support
+- ✅ Provides views for latest metrics and high-risk devices
+
+## Quick Start
+
+### 1. Install Dependencies
+
+```bash
+python3 -m venv venv
+source venv/bin/activate
+pip install -r requirements.txt
+```
+
+### 2. Configure Environment
+
+```bash
+cp .env.example .env
+# Edit .env with your actual values
+```
+
+Required configuration:
+- `INFLUXDB_URL`: Your Scrutiny InfluxDB URL (e.g., `http://192.168.1.100:8086`)
+- `INFLUXDB_TOKEN`: Your Scrutiny InfluxDB token
+- `POSTGRES_*`: Your PostgreSQL connection details
+
+### 3. Set Up PostgreSQL Schema
+
+```bash
+psql -h <host> -U <user> -d <database> -f schema.sql
+```
+
+### 4. Explore Scrutiny's InfluxDB Schema
+
+**IMPORTANT**: Run this first to understand how Scrutiny stores SMART data:
+
+```bash
+python explore_influx.py
+```
+
+This will show you:
+- Available measurements in the `metrics` bucket
+- Field names for each measurement
+- Sample data structure
+- Tags used for device identification
+
+### 5. Customize the Logger
+
+Based on the output from `explore_influx.py`, update `smart_logger.py`:
+
+1. **Update the InfluxDB query** (`fetch_smart_data` method):
+   - Set correct `_measurement` name
+   - Map field names to device properties (temp, serial, model, etc.)
+   - Map SMART attribute fields to IDs (5, 187, 188, 197, 198)
+
+2. **Adjust device detection**:
+   - Update how device path, serial, and model are extracted
+   - Add disk role mapping if you want to match SnapRAID roles
+
+### 6. Run the Logger
+
+```bash
+python smart_logger.py
+```
+
+## Output Format
+
+The logger will display a summary like SnapRAID:
+
+```
+Device: /dev/sda
+  Serial: ZX20AAMV
+  Model: WDC WD200EFAX-68FB5N0
+  Temp: 25°C
+  Power-On: 636 days
+  Failure Probability: 49.00%
+
+================================================================================
+Summary:
+================================================================================
+🔴 HIGH RISK /dev/sdc: 81.00% failure probability
+🟢 OK /dev/sdb: 22.00% failure probability
+```
+
+## PostgreSQL Schema
+
+### Tables
+
+- **`smart_metrics`**: Main table storing all SMART data snapshots
+  - Includes temperature, power-on time, error counts
+  - Stores calculated failure probability
+  - Raw SMART attributes as JSONB for flexibility
+
+### Views
+
+- **`smart_latest`**: Latest metrics for each device
+- **`smart_high_risk`**: Devices with >50% annual failure probability
+
+### Example Queries
+
+```sql
+-- Get latest status for all devices
+SELECT device_path, serial_number, temperature_celsius,
+       failure_probability_pct, timestamp
+FROM smart_latest
+ORDER BY failure_probability_pct DESC;
+
+-- Historical trend for a specific disk
+SELECT timestamp, temperature_celsius, failure_probability_pct
+FROM smart_metrics
+WHERE serial_number = 'ZX20AAMV'
+ORDER BY timestamp DESC
+LIMIT 100;
+
+-- Check specific SMART attribute over time
+SELECT timestamp,
+       (smart_attributes->>'5')::int as reallocated_sectors,
+       (smart_attributes->>'197')::int as pending_sectors
+FROM smart_metrics
+WHERE device_path = '/dev/sda'
+ORDER BY timestamp DESC;
+```
+
+## BackBlaze Algorithm
+
+The failure probability calculation is based on SnapRAID's implementation, which uses BackBlaze's 2014 dataset (47,322 disk observations) to create lookup tables for these SMART attributes:
+
+- **Attribute 5**: Reallocated Sectors Count
+- **Attribute 187**: Reported Uncorrectable Errors
+- **Attribute 188**: Command Timeout (excluded for Seagate disks)
+- **Attribute 197**: Current Pending Sector Count
+- **Attribute 198**: Offline Uncorrectable Sector Count
+
+The algorithm returns the **maximum** AFR across all attributes (since they're correlated, not independent).
+
+## Automation
+
+To run this periodically, add a cron job:
+
+```bash
+# Run every hour
+0 * * * * /path/to/venv/bin/python /path/to/smart_logger.py >> /var/log/smart_logger.log 2>&1
+```
+
+Or use systemd timer, or your preferred scheduler.
+
+## Troubleshooting
+
+### InfluxDB Connection Issues
+
+- Verify the URL and port are correct
+- Check that the token has read permissions for the `metrics` bucket
+- Ensure InfluxDB is accessible from the machine running this script
+
+### No Data Returned
+
+- Run `explore_influx.py` to check the data structure
+- Verify Scrutiny is actively collecting data
+- Check the time range in the query (`-5m` might be too short)
+
+### PostgreSQL Insert Failures
+
+- Ensure the schema was created (`schema.sql`)
+- Check that the database user has INSERT permissions
+- Verify data types match (especially JSONB for smart_attributes)
+
+## License
+
+MIT License - feel free to modify and use as needed!
+
+## Credits
+
+- BackBlaze failure probability tables extracted from [SnapRAID](https://github.com/amadvance/snapraid)
+- Based on BackBlaze's 2014 hard drive reliability dataset