New MQTT adaption

Dockerfile

@@ -0,0 +1,37 @@
+FROM python:3.11-slim
+
+# Set metadata
+LABEL maintainer="mail@hendrikschutter.com"
+LABEL description="Prometheus exporter for VEGAPULS Air sensors via The Things Network"
+LABEL version="2.0"
+
+# Create app directory
+WORKDIR /app
+
+# Install dependencies
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy application files
+COPY ttn-vegapuls-exporter.py .
+COPY config.py .
+
+# Create non-root user
+RUN useradd -r -u 1000 -g users exporter && \
+    chown -R exporter:users /app
+
+# Switch to non-root user
+USER exporter
+
+# Expose metrics port
+EXPOSE 9106
+
+# Health check
+HEALTHCHECK --interval=30s --timeout=10s --start-period=10s --retries=3 \
+    CMD python -c 'import urllib.request; urllib.request.urlopen("http://localhost:9106/health")' || exit 1
+
+# Set environment variables
+ENV PYTHONUNBUFFERED=1
+
+# Run the exporter
+CMD ["python", "ttn-vegapuls-exporter.py"]

README.md

@@ -1,15 +1,232 @@
-# The Things Network Exporter for VEGAPULS Air
+# TTN VEGAPULS Air Prometheus Exporter
 
-Export metrics of a VEGAPULS Air connected via TTN as a prometheus service.
+A robust Prometheus exporter for VEGAPULS Air sensors connected via The Things Network (TTN). This exporter provides reliable monitoring with automatic reconnection, uplink caching, and timeout detection.
 
-## Install ##
+## Features
 
-- `zypper install python311-paho-mqtt`
+- **Uplink Caching**: Stores historical data with timestamps for each device
-- `mkdir /opt/ttn-vegapulsair-exporter/`
+- **Timeout Detection**: Automatically detects offline sensors (configurable, default 19 hours)
-- `cd /opt/ttn-vegapulsair-exporter/`
+- **Better Error Handling**: Comprehensive logging and error recovery
-- import `ttn-vegapulsair-exporter.py` and `config.py`
+- **Multiple Device Support**: Automatically handles multiple sensors
-- Set the constants in `config.py`
+
-- `chmod +x /opt/ttn-vegapulsair-exporter/ttn-vegapulsair-exporter.py` 
+## Metrics Exported
-- `chown -R prometheus /opt/ttn-vegapulsair-exporter/`
+
-- `nano /etc/systemd/system/ttn-vegapulsair-exporter.service`
+### Exporter Metrics
-- `systemctl daemon-reload && systemctl enable --now ttn-vegapulsair-exporter.service`
+- `vegapulsair_exporter_uptime_seconds` - Exporter uptime in seconds
+- `vegapulsair_exporter_requests_total` - Total number of metrics requests
+- `vegapulsair_devices_total` - Total number of known devices
+- `vegapulsair_devices_online` - Number of currently online devices
+
+### Per-Device Metrics
+All device metrics include a `device_id` label:
+
+#### Status Metrics
+- `vegapulsair_device_online{device_id="..."}` - Device online status (1=online, 0=offline)
+- `vegapulsair_last_uplink_seconds_ago{device_id="..."}` - Seconds since last uplink
+
+#### Sensor Measurements
+- `vegapulsair_distance_mm{device_id="..."}` - Distance measurement in millimeters
+- `vegapulsair_temperature_celsius{device_id="..."}` - Temperature in Celsius
+- `vegapulsair_inclination_degrees{device_id="..."}` - Inclination in degrees
+- `vegapulsair_linear_percent{device_id="..."}` - Linear percentage
+- `vegapulsair_percent{device_id="..."}` - Percentage value
+- `vegapulsair_scaled_value{device_id="..."}` - Scaled measurement value
+- `vegapulsair_battery_percent{device_id="..."}` - Remaining battery percentage
+
+#### LoRaWAN Metadata
+- `vegapulsair_rssi_dbm{device_id="..."}` - RSSI in dBm
+- `vegapulsair_channel_rssi_dbm{device_id="..."}` - Channel RSSI in dBm
+- `vegapulsair_snr_db{device_id="..."}` - Signal-to-Noise Ratio in dB
+
+## Requirements
+
+- Python 3.7 or higher
+- `paho-mqtt` library
+
+## Installation
+
+### Option 1: Manual Installation
+
+1. **Install Python dependencies:**
+   ```bash
+   pip install paho-mqtt --break-system-packages
+   # Or use a virtual environment:
+   python3 -m venv venv
+   source venv/bin/activate
+   pip install paho-mqtt
+   ```
+
+2. **Create installation directory:**
+   ```bash
+   sudo mkdir -p /opt/ttn-vegapuls-exporter
+   cd /opt/ttn-vegapuls-exporter
+   ```
+
+3. **Copy files:**
+   ```bash
+   sudo cp ttn-vegapuls-exporter.py /opt/ttn-vegapuls-exporter/
+   sudo cp config.py /opt/ttn-vegapuls-exporter/
+   sudo chmod +x /opt/ttn-vegapuls-exporter/ttn-vegapuls-exporter.py
+   ```
+
+4. **Configure the exporter:**
+   ```bash
+   sudo nano /opt/ttn-vegapuls-exporter/config.py
+   ```
+   
+   Set the following required parameters:
+   - `ttn_user`: Your TTN application ID (format: `your-app-id@ttn`)
+   - `ttn_key`: Your TTN API key (get from TTN Console)
+   - `ttn_region`: Your TTN region (EU1, NAM1, AU1, etc.)
+
+5. **Set permissions:**
+   ```bash
+   sudo useradd -r -s /bin/false prometheus  # If user doesn't exist
+   sudo chown -R prometheus:prometheus /opt/ttn-vegapuls-exporter
+   ```
+
+6. **Install systemd service:**
+   ```bash
+   sudo cp ttn-vegapuls-exporter.service /etc/systemd/system/
+   sudo systemctl daemon-reload
+   sudo systemctl enable ttn-vegapuls-exporter.service
+   sudo systemctl start ttn-vegapuls-exporter.service
+   ```
+
+7. **Check status:**
+   ```bash
+   sudo systemctl status ttn-vegapuls-exporter.service
+   sudo journalctl -u ttn-vegapuls-exporter.service -f
+   ```
+
+### Option 2: Docker Installation
+See `docker-compose.yml`.
+
+## Configuration
+
+Edit `config.py` to customize the exporter:
+
+```python
+# HTTP Server configuration
+hostName = "0.0.0.0"  # Listen address
+serverPort = 9106      # Port for metrics endpoint
+
+# TTN Configuration
+ttn_user = "your-app@ttn"
+ttn_key = "NNSXS...."  # From TTN Console
+ttn_region = "EU1"
+
+# Timeout configuration
+sensor_timeout_hours = 19  # Mark sensor offline after N hours
+
+# Logging
+log_level = "INFO"  # DEBUG, INFO, WARNING, ERROR, CRITICAL
+```
+
+### Getting TTN Credentials
+
+1. Log in to [TTN Console](https://console.cloud.thethings.network/)
+2. Select your application
+3. Go to **Integrations** → **MQTT**
+4. Copy the following:
+   - **Username**: Your application ID (format: `your-app-id@ttn`)
+   - **Password**: Generate an API key with "Read application traffic" permission
+   - **Region**: Your cluster region (visible in the URL, e.g., `eu1`)
+
+## Prometheus Configuration
+
+Add to your `prometheus.yml`:
+
+```yaml
+scrape_configs:
+  - job_name: 'vegapuls-air'
+    static_configs:
+      - targets: ['localhost:9106']
+    scrape_interval: 60s
+    scrape_timeout: 10s
+```
+
+### Example Prometheus Alerts
+
+See `prometheus-alerts.yml`.
+
+## Troubleshooting
+
+### No Metrics Appearing
+
+1. **Check MQTT connection:**
+   ```bash
+   sudo journalctl -u ttn-vegapuls-exporter.service | grep MQTT
+   ```
+   
+   You should see: `Successfully connected to TTN MQTT broker`
+
+2. **Verify TTN credentials:**
+   - Ensure `ttn_user` format is correct: `your-app-id@ttn`
+   - Verify API key has "Read application traffic" permission
+   - Check region matches your TTN cluster
+
+3. **Test metrics endpoint:**
+   ```bash
+   curl http://localhost:9106/metrics
+   ```
+
+### MQTT Disconnections
+
+The exporter now handles disconnections automatically with exponential backoff. Check logs:
+
+```bash
+sudo journalctl -u ttn-vegapuls-exporter.service -f
+```
+
+If disconnections persist:
+- Check network connectivity to TTN
+- Verify firewall allows outbound port 8883
+- Ensure system time is correct (TLS certificates)
+
+### Devices Not Appearing
+
+1. **Verify devices are sending uplinks:**
+   - Check TTN Console → Applications → Your App → Live Data
+   - Ensure devices are joined and transmitting
+
+2. **Check user ID:**
+   - `ttn_user` must match your TTN application ID exactly
+
+3. **Verify payload decoder:**
+   - Devices must have decoded payload in TTN
+   - Check TTN Payload Formatter is configured
+
+### Debug Mode
+
+Enable debug logging in `config.py`:
+
+```python
+log_level = "DEBUG"
+```
+
+This will show:
+- All MQTT messages received
+- Cache updates
+- Device status changes
+- Detailed error information
+
+### Data Flow
+
+```
+VEGAPULS Air Sensor
+    ↓
+LoRaWAN Gateway
+    ↓
+The Things Network
+    ↓
+MQTT Broker (TLS)
+    ↓
+Exporter (caches data)
+    ↓
+Prometheus (scrapes metrics)
+```
+
+## License
+
+See [LICENSE](LICENSE) file for details.

config.py

@@ -1,12 +1,39 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
-""" Author:                     Hendrik Schutter, mail@hendrikschutter.com
+"""
+Configuration for TTN VEGAPULS Air Prometheus Exporter
+Author: Hendrik Schutter, mail@hendrikschutter.com
 """
 
+# HTTP Server configuration
 hostName = "127.0.0.1"
 serverPort = 9106
 exporter_prefix = "vegapulsair_"
 
-ttn_user = "appid@ttn"
+# TTN MQTT Configuration
-ttn_key = "THE APP API KEY FROM TTN CONSOLE"
+# Get your credentials from TTN Console -> Applications -> Your App -> Integrations -> MQTT
-ttn_region = "EU1"	
+ttn_user = "appid@ttn"  # Your application ID
+ttn_key = "THE APP API KEY FROM TTN CONSOLE"  # Your API key
+ttn_region = "EU1"  # TTN region: EU1, NAM1, AU1, etc.
+
+# Integration method: "mqtt" or "http"
+# - mqtt: Subscribe to TTN MQTT broker (recommended for real-time updates)
+# - http: Use HTTP Integration webhook (requires TTN webhook configuration)
+integration_method = "mqtt"
+
+# Timeout configuration
+# Time in hours after which a sensor is considered offline if no uplink is received
+sensor_timeout_hours = 19
+
+# MQTT specific settings
+mqtt_keepalive = 60  # MQTT keepalive interval in seconds
+mqtt_reconnect_delay = 5  # Delay between reconnection attempts in seconds
+mqtt_reconnect_max_delay = 300  # Maximum delay between reconnection attempts
+
+# Logging configuration
+log_level = "INFO"  # DEBUG, INFO, WARNING, ERROR, CRITICAL
+log_format = "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
+
+# Cache configuration
+cache_cleanup_interval = 3600  # Cleanup old cache entries every hour
+max_cache_age_hours = 72  # Remove cache entries older than 72 hours

docker-compose.yml

@@ -0,0 +1,65 @@
+version: '3.8'
+
+services:
+  ttn-vegapuls-exporter:
+    image: python:3.11-slim
+    container_name: ttn-vegapuls-exporter
+    restart: unless-stopped
+    
+    # Install dependencies and run exporter
+    entrypoint: |
+      sh -c "pip install --no-cache-dir paho-mqtt && python ttn-vegapuls-exporter.py"
+    
+    working_dir: /app
+    
+    # Expose metrics port
+    ports:
+      - "9106:9106"
+    
+    # Mount application files (read-only)
+    volumes:
+      - ./ttn-vegapuls-exporter.py:/app/ttn-vegapuls-exporter.py:ro
+      - ./config.py:/app/config.py:ro
+    
+    # Environment variables
+    environment:
+      - PYTHONUNBUFFERED=1
+    
+    # Health check
+    healthcheck:
+      test: ["CMD-SHELL", "python -c 'import urllib.request; urllib.request.urlopen(\"http://localhost:9106/health\")' || exit 1"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 10s
+    
+    # Resource limits
+    deploy:
+      resources:
+        limits:
+          memory: 256M
+          cpus: '0.05'
+        reservations:
+          memory: 64M
+    
+    # Logging configuration
+    logging:
+      driver: "json-file"
+      options:
+        max-size: "10m"
+        max-file: "3"
+    
+    # Network configuration
+    networks:
+      - monitoring
+    
+    # Security options
+    security_opt:
+      - no-new-privileges:true
+    
+    # Run as non-root user
+    user: "1000:1000"
+
+networks:
+  monitoring:
+    driver: bridge

prometheus-alerts.yml

@@ -0,0 +1,204 @@
+# Prometheus Alert Rules for VEGAPULS Air Sensors
+#
+# Installation:
+# 1. Copy this file to /etc/prometheus/rules/vegapuls-alerts.yml
+# 2. Add to prometheus.yml:
+#    rule_files:
+#      - /etc/prometheus/rules/vegapuls-alerts.yml
+# 3. Reload Prometheus: systemctl reload prometheus
+
+groups:
+  - name: ttn_vegapuls_air_alerts
+    interval: 60s
+    rules:
+      # === Exporter Health ===
+      
+      - alert: VEGAPULSExporterDown
+        expr: up{job="vegapuls-air"} == 0
+        for: 5m
+        labels:
+          severity: critical
+          component: exporter
+        annotations:
+          summary: "VEGAPULS Air exporter is down"
+          description: "The VEGAPULS Air Prometheus exporter has been down for more than 5 minutes. Check the service status."
+          runbook: "Check systemctl status vegapuls-exporter and journalctl -u vegapuls-exporter"
+      
+      # === Device Online Status ===
+      
+      - alert: VEGAPULSSensorOffline
+        expr: vegapulsair_device_online == 0
+        for: 10m
+        labels:
+          severity: warning
+          component: sensor
+        annotations:
+          summary: "VEGAPULS sensor {{ $labels.device_id }} is offline"
+          description: "Sensor {{ $labels.device_id }} has not sent an uplink for more than 19 hours and is considered offline."
+          runbook: "Check sensor battery, LoRaWAN coverage, and TTN Console for error messages"
+      
+      - alert: VEGAPULSSensorMissing
+        expr: |
+          (time() - vegapulsair_last_uplink_seconds_ago) > 86400
+        for: 30m
+        labels:
+          severity: critical
+          component: sensor
+        annotations:
+          summary: "VEGAPULS sensor {{ $labels.device_id }} missing for over 24h"
+          description: "Sensor {{ $labels.device_id }} has not transmitted for over 24 hours. Last uplink: {{ $value | humanizeDuration }} ago."
+          runbook: "Physical inspection required. Check sensor power and installation."
+      
+      # === Battery Monitoring ===
+      
+      - alert: VEGAPULSBatteryCritical
+        expr: vegapulsair_battery_percent < 10
+        for: 1h
+        labels:
+          severity: critical
+          component: battery
+        annotations:
+          summary: "VEGAPULS sensor {{ $labels.device_id }} battery critically low"
+          description: "Battery level at {{ $value }}%. Sensor will stop functioning soon. Immediate replacement required."
+          runbook: "Schedule urgent battery replacement"
+      
+      - alert: VEGAPULSBatteryLow
+        expr: vegapulsair_battery_percent < 20
+        for: 6h
+        labels:
+          severity: warning
+          component: battery
+        annotations:
+          summary: "VEGAPULS sensor {{ $labels.device_id }} battery low"
+          description: "Battery level at {{ $value }}%. Plan battery replacement soon."
+          runbook: "Schedule battery replacement within 2-4 weeks"
+      
+      - alert: VEGAPULSBatteryWarning
+        expr: vegapulsair_battery_percent < 30
+        for: 12h
+        labels:
+          severity: info
+          component: battery
+        annotations:
+          summary: "VEGAPULS sensor {{ $labels.device_id }} battery below 30%"
+          description: "Battery level at {{ $value }}%. Monitor and plan replacement."
+          runbook: "Add to maintenance schedule for next quarter"
+      
+      # === Signal Quality ===
+      
+      - alert: VEGAPULSWeakSignal
+        expr: vegapulsair_rssi_dbm < -120
+        for: 1h
+        labels:
+          severity: warning
+          component: network
+        annotations:
+          summary: "VEGAPULS sensor {{ $labels.device_id }} has weak signal"
+          description: "RSSI is {{ $value }} dBm (very weak). May indicate coverage issues or antenna problems."
+          runbook: "Check gateway coverage, sensor placement, and antenna connection"
+      
+      - alert: VEGAPULSPoorSNR
+        expr: vegapulsair_snr_db < -15
+        for: 1h
+        labels:
+          severity: warning
+          component: network
+        annotations:
+          summary: "VEGAPULS sensor {{ $labels.device_id }} has poor SNR"
+          description: "Signal-to-Noise Ratio is {{ $value }} dB. Signal quality is degraded."
+          runbook: "Check for interference, gateway issues, or repositioning sensor"
+      
+      # === Temperature Monitoring ===
+      
+      - alert: VEGAPULSTemperatureExtreme
+        expr: |
+          vegapulsair_temperature_celsius > 60 or
+          vegapulsair_temperature_celsius < -20
+        for: 30m
+        labels:
+          severity: warning
+          component: environment
+        annotations:
+          summary: "VEGAPULS sensor {{ $labels.device_id }} extreme temperature"
+          description: "Temperature is {{ $value }}°C, outside normal operating range."
+          runbook: "Check sensor location and environmental conditions"
+      
+      # === Data Quality ===
+      
+      - alert: VEGAPULSNoDataReceived
+        expr: |
+          rate(vegapulsair_exporter_requests_total[5m]) > 0 and
+          vegapulsair_devices_total == 0
+        for: 15m
+        labels:
+          severity: warning
+          component: integration
+        annotations:
+          summary: "VEGAPULS exporter receiving no device data"
+          description: "Exporter is running and being scraped, but no device data is available. Check MQTT connection and TTN configuration."
+          runbook: "Check exporter logs, TTN Console live data, and MQTT credentials"
+      
+      - alert: VEGAPULSAllDevicesOffline
+        expr: |
+          vegapulsair_devices_total > 0 and
+          vegapulsair_devices_online == 0
+        for: 30m
+        labels:
+          severity: critical
+          component: system
+        annotations:
+          summary: "All VEGAPULS sensors are offline"
+          description: "{{ $value }} devices are registered but none are online. System-wide issue suspected."
+          runbook: "Check TTN gateway status, network connectivity, and power supply"
+      
+      # === Performance Monitoring ===
+      
+      - alert: VEGAPULSHighScrapeRate
+        expr: rate(vegapulsair_exporter_requests_total[5m]) > 2
+        for: 10m
+        labels:
+          severity: info
+          component: performance
+        annotations:
+          summary: "High scrape rate on VEGAPULS exporter"
+          description: "Prometheus is scraping at {{ $value }} requests/second. Consider increasing scrape_interval."
+          runbook: "Review Prometheus configuration and adjust scrape_interval if needed"
+
+# === Recording Rules for Easier Querying ===
+
+  - name: vegapuls_air_recording_rules
+    interval: 60s
+    rules:
+      # Battery drain rate (percent per day)
+      - record: vegapulsair_battery_drain_rate_percent_per_day
+        expr: |
+          rate(vegapulsair_battery_percent[7d]) * -86400
+      
+      # Average signal strength per device (7 day)
+      - record: vegapulsair_rssi_avg_7d
+        expr: |
+          avg_over_time(vegapulsair_rssi_dbm[7d])
+      
+      # Uplink frequency (uplinks per day)
+      - record: vegapulsair_uplink_frequency_per_day
+        expr: |
+          86400 / avg_over_time(vegapulsair_last_uplink_seconds_ago[7d])
+      
+      # Device availability percentage (24h)
+      - record: vegapulsair_device_availability_percent_24h
+        expr: |
+          avg_over_time(vegapulsair_device_online[24h]) * 100
+
+# === Usage Examples ===
+#
+# Query battery drain rate:
+#   vegapulsair_battery_drain_rate_percent_per_day
+#
+# Query devices with availability < 95%:
+#   vegapulsair_device_availability_percent_24h < 95
+#
+# Query average RSSI over 7 days:
+#   vegapulsair_rssi_avg_7d
+#
+# Query uplink frequency:
+#   vegapulsair_uplink_frequency_per_day

requirements.txt

@@ -0,0 +1,4 @@
+# TTN VEGAPULS Air Exporter - Python Dependencies
+
+# MQTT client for connecting to The Things Network
+paho-mqtt>=2.0.0,<3.0.0

ttn-vegapulsair-exporter.py

@@ -1,286 +1,595 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
-""" Author:                     Hendrik Schutter, mail@hendrikschutter.com
+"""
+TTN VEGAPULS Air Prometheus Exporter
+Exports metrics from VEGAPULS Air sensors connected via The Things Network
+
+Author: Hendrik Schutter, mail@hendrikschutter.com
 """
 
-from http.server import BaseHTTPRequestHandler, HTTPServer
-import paho.mqtt.client as mqtt
-from datetime import datetime, timedelta
-import threading
-import time
-import json
 import sys
-import config
+import json
+import time
+import threading
 import logging
 import ssl
+from datetime import datetime, timedelta
+from http.server import BaseHTTPRequestHandler, HTTPServer
+from typing import Dict, Optional, Any
+
+import paho.mqtt.client as mqtt
+
+import config
 
 
-scrape_healthy = True
+class SensorDataCache:
-startTime = datetime.now()
+    """Thread-safe cache for sensor uplink data with timeout tracking"""
-lastMqttReception = datetime.now()
-node_metrics = list()
-mutex = threading.Lock()
-request_count = 0
 
-mqtt_client = None
+    def __init__(self, timeout_hours: int = 19):
-mqtt_connected = False
+        self._data: Dict[str, Dict[str, Any]] = {}
-mqtt_lock = threading.Lock()
+        self._lock = threading.RLock()
+        self.timeout_hours = timeout_hours
 
-def monitor_timeout():
+    def update(
-    global scrape_healthy
+        self, device_id: str, payload: Dict, metadata: list, timestamp: datetime
-    global lastMqttReception
+    ):
-    global mqtt_connected
+        """
+        Update cached data for a device
 
-    while True:
+        Args:
-        time_since_last_reception = datetime.now() - lastMqttReception
+            device_id: Unique device identifier
-        if time_since_last_reception > timedelta(hours=config.ttn_timeout):
+            payload: Decoded payload from TTN
-            with mutex:
+            metadata: RX metadata from TTN
-                scrape_healthy = False
+            timestamp: Timestamp of the uplink
-                mqtt_connected = False
+        """
-        time.sleep(60)  # Check timeout every minute
+        with self._lock:
+            self._data[device_id] = {
+                "payload": payload,
+                "metadata": metadata,
+                "timestamp": timestamp,
+                "is_online": True,
+            }
+            logging.info(f"Updated cache for device {device_id}")
 
-def reconnect_mqtt():
+    def get_all_devices(self) -> Dict[str, Dict[str, Any]]:
-    global mqtt_client
+        """
-    global mqtt_connected
+        Get all cached device data
 
-    while True:
+        Returns:
-        if not mqtt_connected:
+            Dictionary of device data
-            with mqtt_lock:
+        """
-                try:
+        with self._lock:
-                    if mqtt_client is None:
+            return dict(self._data)
-                        print("MQTT client is None, creating a new client...")
-                        mqtt_client = mqtt.Client(mqtt.CallbackAPIVersion.VERSION2)
-                        mqtt_client.on_connect = on_connect
-                        mqtt_client.on_message = on_message
-                        mqtt_client.on_disconnect = on_disconnect
-                        mqtt_client.username_pw_set(config.ttn_user, config.ttn_key)
-                        mqtt_client.tls_set()
 
-                    print("Attempting to reconnect to MQTT broker...")
+    def check_timeouts(self):
-                    mqtt_client.connect(
+        """Check all devices for timeout and mark offline ones"""
-                        config.ttn_region.lower() + ".cloud.thethings.network", 8883, 60
+        with self._lock:
+            now = datetime.now()
+            timeout_threshold = timedelta(hours=self.timeout_hours)
+
+            for device_id, data in self._data.items():
+                time_since_update = now - data["timestamp"]
+                was_online = data["is_online"]
+                data["is_online"] = time_since_update < timeout_threshold
+
+                if was_online and not data["is_online"]:
+                    logging.warning(
+                        f"Device {device_id} marked as OFFLINE "
+                        f"(no uplink for {time_since_update.total_seconds()/3600:.1f} hours)"
                     )
-                except Exception as e:
+                elif not was_online and data["is_online"]:
-                    print(f"MQTT reconnect failed: {e}")
+                    logging.info(f"Device {device_id} is back ONLINE")
-        time.sleep(60)  # Retry every 10 seconds
+
+    def cleanup_old_entries(self, max_age_hours: int = 72):
+        """Remove entries older than max_age_hours"""
+        with self._lock:
+            now = datetime.now()
+            max_age = timedelta(hours=max_age_hours)
+
+            devices_to_remove = [
+                device_id
+                for device_id, data in self._data.items()
+                if now - data["timestamp"] > max_age
+            ]
+
+            for device_id in devices_to_remove:
+                del self._data[device_id]
+                logging.info(f"Removed stale cache entry for device {device_id}")
 
 
-class RequestHandler(BaseHTTPRequestHandler):
+class TTNMQTTClient:
-    def log_message(self, format, *args):
+    """Manages MQTT connection to TTN with automatic reconnection"""
-        pass
 
-    def get_metrics(self):
+    def __init__(self, cache: SensorDataCache, config_module):
-        global request_count
+        self.cache = cache
-        global node_metrics
+        self.config = config_module
-        global mutex
+        self.client: Optional[mqtt.Client] = None
-        mutex.acquire()
+        self.connected = False
-        self.send_response(200)
+        self._lock = threading.Lock()
-        self.send_header("Content-type", "text/html")
+        self._should_run = True
-        self.end_headers()
-        self.wfile.write(
-            bytes(
-                config.exporter_prefix
-                + "exporter_duration_seconds_sum "
-                + str(int((datetime.now() - startTime).total_seconds()))
-                + "\n",
-                "utf-8",
-            )
-        )
-        self.wfile.write(
-            bytes(
-                config.exporter_prefix
-                + "exporter_request_count "
-                + str(request_count)
-                + "\n",
-                "utf-8",
-            )
-        )
-        self.wfile.write(
-            bytes(
-                config.exporter_prefix
-                + "exporter_scrape_healthy "
-                + str(int(scrape_healthy))
-                + "\n",
-                "utf-8",
-            )
-        )
 
-        for metric in node_metrics:
+        # Setup logging
-            self.wfile.write(bytes(config.exporter_prefix + metric + "\n", "utf-8"))
+        self.logger = logging.getLogger("TTNMQTTClient")
 
-        mutex.release()
+    def _on_connect(self, client, userdata, flags, reason_code, properties):
+        """Callback when connected to MQTT broker"""
+        if reason_code == 0:
+            self.logger.info("Successfully connected to TTN MQTT broker")
+            self.connected = True
 
-    def do_GET(self):
+            # Subscribe to uplink messages
-        global request_count
+            topic = f"v3/{self.config.ttn_user}/devices/+/up"
-        request_count += 1
+            client.subscribe(topic, qos=1)
-        if self.path.startswith("/metrics"):
+            self.logger.info(f"Subscribed to topic: {topic}")
-            self.get_metrics()
         else:
-            self.send_response(200)
+            self.logger.error(
-            self.send_header("Content-type", "text/html")
+                f"Failed to connect to MQTT broker. Reason code: {reason_code}"
-            self.end_headers()
-            self.wfile.write(bytes("<html>", "utf-8"))
-            self.wfile.write(
-                bytes("<head><title>VEGAPULS Air exporter</title></head>", "utf-8")
             )
-            self.wfile.write(bytes("<body>", "utf-8"))
+            self.connected = False
-            self.wfile.write(
+
-                bytes(
+    def _on_disconnect(self, client, userdata, flags, reason_code, properties):
-                    "<h1>ttn-vegapulsair exporter based on data from LoRaWAN TTN node.</h1>",
+        """Callback when disconnected from MQTT broker"""
-                    "utf-8",
+        self.logger.warning(
+            f"Disconnected from MQTT broker. Reason code: {reason_code}"
+        )
+        self.connected = False
+
+    def _on_message(self, client, userdata, msg):
+        """Callback when a message is received"""
+        self.logger.debug(f"Uplink message received! {msg.topic}")
+        try:
+            # Parse the JSON payload
+            message_data = json.loads(msg.payload.decode("utf-8"))
+
+            # Extract device information
+            device_id = message_data.get("end_device_ids", {}).get(
+                "device_id", "unknown"
+            )
+
+            # Check if this is an uplink message with decoded payload
+            if "uplink_message" not in message_data:
+                self.logger.debug(f"Ignoring non-uplink message from {device_id}")
+                return
+
+            uplink = message_data["uplink_message"]
+
+            if "decoded_payload" not in uplink:
+                self.logger.warning(f"No decoded payload for device {device_id}")
+                return
+
+            # Update cache with new data
+            self.cache.update(
+                device_id=device_id,
+                payload=uplink["decoded_payload"],
+                metadata=uplink.get("rx_metadata", []),
+                timestamp=datetime.now(),
+            )
+
+            self.logger.debug(f"Processed uplink from device: {device_id}")
+
+        except json.JSONDecodeError as e:
+            self.logger.error(f"Failed to parse MQTT message: {e}")
+        except Exception as e:
+            self.logger.error(f"Error processing MQTT message: {e}", exc_info=True)
+
+    def _create_client(self):
+        """Create and configure MQTT client"""
+        client = mqtt.Client(
+            client_id=f"vegapuls-exporter-{int(time.time())}",
+            callback_api_version=mqtt.CallbackAPIVersion.VERSION2,
+        )
+
+        # Set callbacks
+        client.on_connect = self._on_connect
+        client.on_disconnect = self._on_disconnect
+        client.on_message = self._on_message
+
+        # Set credentials
+        client.username_pw_set(self.config.ttn_user, self.config.ttn_key)
+
+        # Configure TLS
+        client.tls_set(cert_reqs=ssl.CERT_REQUIRED, tls_version=ssl.PROTOCOL_TLS_CLIENT)
+        client.tls_insecure_set(False)
+
+        return client
+
+    def connect(self):
+        """Connect to TTN MQTT broker"""
+        with self._lock:
+            try:
+                if self.client is None:
+                    self.client = self._create_client()
+
+                broker_url = f"{self.config.ttn_region.lower()}.cloud.thethings.network"
+                self.logger.info(f"Connecting to MQTT broker: {broker_url}")
+
+                self.client.connect(
+                    broker_url, port=8883, keepalive=self.config.mqtt_keepalive
+                )
+
+                # Start the network loop in a separate thread
+                self.client.loop_start()
+
+                return True
+
+            except Exception as e:
+                self.logger.error(f"Failed to connect to MQTT broker: {e}")
+                return False
+
+    def disconnect(self):
+        """Disconnect from MQTT broker"""
+        with self._lock:
+            if self.client:
+                self.client.loop_stop()
+                self.client.disconnect()
+                self.connected = False
+                self.logger.info("Disconnected from MQTT broker")
+
+    def run_with_reconnect(self):
+        """Main loop with automatic reconnection"""
+        reconnect_delay = self.config.mqtt_reconnect_delay
+
+        while self._should_run:
+            if not self.connected:
+                self.logger.info("Attempting to connect to MQTT broker...")
+
+                if self.connect():
+                    # Reset reconnect delay on successful connection
+                    reconnect_delay = self.config.mqtt_reconnect_delay
+                else:
+                    # Exponential backoff for reconnection
+                    self.logger.warning(
+                        f"Reconnection failed. Retrying in {reconnect_delay}s..."
+                    )
+                    time.sleep(reconnect_delay)
+                    reconnect_delay = min(
+                        reconnect_delay * 2, self.config.mqtt_reconnect_max_delay
+                    )
+                    continue
+
+            # Wait a bit before checking connection again
+            time.sleep(10)
+
+    def stop(self):
+        """Stop the MQTT client"""
+        self._should_run = False
+        self.disconnect()
+
+
+class MetricsServer:
+    """HTTP server for Prometheus metrics endpoint"""
+
+    def __init__(self, cache: SensorDataCache, config_module):
+        self.cache = cache
+        self.config = config_module
+        self.start_time = datetime.now()
+        self.request_count = 0
+        self._lock = threading.Lock()
+
+    def _format_metric(
+        self, name: str, value: Any, labels: Dict[str, str] = None
+    ) -> str:
+        """
+        Format a Prometheus metric
+
+        Args:
+            name: Metric name
+            value: Metric value
+            labels: Optional labels dictionary
+
+        Returns:
+            Formatted metric string
+        """
+        metric_name = f"{self.config.exporter_prefix}{name}"
+
+        if labels:
+            label_str = ",".join([f'{k}="{v}"' for k, v in labels.items()])
+            return f"{metric_name}{{{label_str}}} {value}"
+        else:
+            return f"{metric_name} {value}"
+
+    def _generate_metrics(self) -> str:
+        """Generate all Prometheus metrics"""
+        metrics = []
+
+        # Exporter meta metrics
+        uptime = int((datetime.now() - self.start_time).total_seconds())
+        metrics.append(self._format_metric("exporter_uptime_seconds", uptime))
+        metrics.append(
+            self._format_metric("exporter_requests_total", self.request_count)
+        )
+
+        # Get all device data
+        devices = self.cache.get_all_devices()
+
+        # Overall health metric
+        online_devices = sum(1 for d in devices.values() if d["is_online"])
+        total_devices = len(devices)
+        metrics.append(self._format_metric("devices_total", total_devices))
+        metrics.append(self._format_metric("devices_online", online_devices))
+
+        # Per-device metrics
+        for device_id, data in devices.items():
+            labels = {"device_id": device_id}
+
+            # Device online status (1 = online, 0 = offline/timeout)
+            metrics.append(
+                self._format_metric("device_online", int(data["is_online"]), labels)
+            )
+
+            # Time since last uplink in seconds
+            time_since_uplink = (datetime.now() - data["timestamp"]).total_seconds()
+            metrics.append(
+                self._format_metric(
+                    "last_uplink_seconds_ago", int(time_since_uplink), labels
                 )
             )
-            self.wfile.write(bytes('<p><a href="/metrics">Metrics</a></p>', "utf-8"))
-            self.wfile.write(bytes("</body>", "utf-8"))
-            self.wfile.write(bytes("</html>", "utf-8"))
 
-def update_metrics(payload, metadata):
+            payload = data["payload"]
-    global node_metrics
+            metadata = data["metadata"]
-    global mutex
-    global scrape_healthy
-    global lastMqttReception
 
-    mutex.acquire()
+            # Sensor measurements
-    node_metrics.clear()
+            if "Distance" in payload:
+                metrics.append(
+                    self._format_metric(
+                        "distance_mm", float(payload["Distance"]), labels
+                    )
+                )
 
-    if "Distance" in payload:
+            if "Temperature" in payload:
-        node_metrics.append("distance " + str(float(payload["Distance"])))
+                metrics.append(
+                    self._format_metric(
+                        "temperature_celsius", int(payload["Temperature"]), labels
+                    )
+                )
 
-    if "Inclination_degree" in payload:
+            if "Inclination_degree" in payload:
-        node_metrics.append("inclination_degree " + str(int(payload["Inclination_degree"])))
+                metrics.append(
+                    self._format_metric(
+                        "inclination_degrees",
+                        int(payload["Inclination_degree"]),
+                        labels,
+                    )
+                )
 
-    if "MvLinProcent" in payload:
+            if "MvLinProcent" in payload:
-        node_metrics.append("linprocent " + str(int(payload["MvLinProcent"])))
+                metrics.append(
+                    self._format_metric(
+                        "linear_percent", int(payload["MvLinProcent"]), labels
+                    )
+                )
 
-    if "MvProcent" in payload:
+            if "MvProcent" in payload:
-        node_metrics.append("procent " + str(int(payload["MvProcent"])))
+                metrics.append(
+                    self._format_metric("percent", int(payload["MvProcent"]), labels)
+                )
 
-    if "MvScaled" in payload:
+            if "MvScaled" in payload:
-        node_metrics.append("scaled " + str(float(payload["MvScaled"])))
+                metrics.append(
+                    self._format_metric(
+                        "scaled_value", float(payload["MvScaled"]), labels
+                    )
+                )
 
-    if "MvScaledUnit" in payload:
+            if "MvScaledUnit" in payload:
-        node_metrics.append("scaled_unit " + str(int(payload["MvScaledUnit"])))
+                metrics.append(
+                    self._format_metric(
+                        "scaled_unit", int(payload["MvScaledUnit"]), labels
+                    )
+                )
 
-    if "PacketIdentifier" in payload:
+            if "PacketIdentifier" in payload:
-        node_metrics.append("packet_identifier " + str(int(payload["PacketIdentifier"])))
+                metrics.append(
+                    self._format_metric(
+                        "packet_identifier", int(payload["PacketIdentifier"]), labels
+                    )
+                )
 
-    if "RemainingPower" in payload:
+            if "RemainingPower" in payload:
-        node_metrics.append("remaining_power " + str(int(payload["RemainingPower"])))
+                metrics.append(
+                    self._format_metric(
+                        "battery_percent", int(payload["RemainingPower"]), labels
+                    )
+                )
 
-    if "Temperature" in payload:
+            if "Unit" in payload:
-        node_metrics.append("temperature " + str(int(payload["Temperature"])))
+                metrics.append(
+                    self._format_metric("unit", int(payload["Unit"]), labels)
+                )
 
-    if "Unit" in payload:
+            if "UnitTemperature" in payload:
-        node_metrics.append("unit " + str(int(payload["Unit"])))
+                metrics.append(
+                    self._format_metric(
+                        "temperature_unit", int(payload["UnitTemperature"]), labels
+                    )
+                )
 
-    if "UnitTemperature" in payload:
+            # LoRaWAN metadata
-        node_metrics.append("temperature_unit " + str(int(payload["UnitTemperature"])))
+            if metadata and len(metadata) > 0:
+                first_gateway = metadata[0]
 
-    if "rssi" in metadata[0]:
+                if "rssi" in first_gateway:
-        node_metrics.append("rssi " + str(int(metadata[0]["rssi"])))
+                    metrics.append(
+                        self._format_metric(
+                            "rssi_dbm", int(first_gateway["rssi"]), labels
+                        )
+                    )
 
-    if "channel_rssi" in metadata[0]:
+                if "channel_rssi" in first_gateway:
-        node_metrics.append("channel_rssi " + str(int(metadata[0]["channel_rssi"])))
+                    metrics.append(
+                        self._format_metric(
+                            "channel_rssi_dbm",
+                            int(first_gateway["channel_rssi"]),
+                            labels,
+                        )
+                    )
 
-    if "snr" in metadata[0]:
+                if "snr" in first_gateway:
-        node_metrics.append("snr " + str(float(metadata[0]["snr"])))
+                    metrics.append(
+                        self._format_metric(
+                            "snr_db", float(first_gateway["snr"]), labels
+                        )
+                    )
 
-    scrape_healthy = True
+        return "\n".join(metrics) + "\n"
-    lastMqttReception = datetime.now()
-    mutex.release()
 
-def on_connect(client, userdata, flags, reason_code, properties):
+    def create_handler(self):
-    global mqtt_connected
+        """Create HTTP request handler"""
-    if reason_code == 0:
+        server_instance = self
-        print("\nConnected to MQTT: reason_code = " + str(reason_code))
-        mqtt_connected = True
-    elif reason_code > 0:
-        print("\nNot connected to MQTT: reason_code = " + str(reason_code))
-        mqtt_connected = False
 
-def on_disconnect(client, userdata, flags, reason_code, tmp):
+        class RequestHandler(BaseHTTPRequestHandler):
-    global mqtt_connected
+            def log_message(self, format, *args):
-    print(f"Disconnected from MQTT: reason_code = {reason_code}")
+                """Suppress default logging"""
-    mqtt_connected = False
+                pass
 
-def on_message(mqttc, obj, msg):
+            def do_GET(self):
-    print("on_message")
+                with server_instance._lock:
-    global scrape_healthy
+                    server_instance.request_count += 1
 
-    try:
+                if self.path == "/metrics":
-        parsedJSON = json.loads(msg.payload)
+                    self.send_response(200)
-        print(parsedJSON)
+                    self.send_header("Content-Type", "text/plain; charset=utf-8")
-        uplink_message = parsedJSON["uplink_message"]
+                    self.end_headers()
-        update_metrics(uplink_message["decoded_payload"], uplink_message["rx_metadata"])
-    except Exception as e:
-        with mutex:
-            scrape_healthy = False
-        print(f"Unable to parse uplink: {e}")
 
-def poll_mqtt(mqtt_client):
+                    metrics = server_instance._generate_metrics()
-    # Start the network loop
+                    self.wfile.write(metrics.encode("utf-8"))
-    mqtt_client.loop_forever()
 
-def configure_mqtt_client():
+                elif self.path == "/" or self.path == "/health":
-    client = mqtt.Client(mqtt.CallbackAPIVersion.VERSION2)
+                    self.send_response(200)
-    client.on_connect = on_connect
+                    self.send_header("Content-Type", "text/html; charset=utf-8")
-    client.on_message = on_message
+                    self.end_headers()
-    client.on_disconnect = on_disconnect
 
-    # Set credentials
+                    html = """
-    client.username_pw_set(config.ttn_user, config.ttn_key)
+                    <html>
+                    <head><title>VEGAPULS Air Exporter</title></head>
+                    <body>
+                        <h1>TTN VEGAPULS Air Prometheus Exporter</h1>
+                        <p>Exporter for VEGAPULS Air sensors connected via The Things Network</p>
+                        <p><a href="/metrics">Metrics</a></p>
+                    </body>
+                    </html>
+                    """
+                    self.wfile.write(html.encode("utf-8"))
 
-    # Set up TLS/SSL
+                else:
-    client.tls_set(
+                    self.send_response(404)
-        cert_reqs=ssl.CERT_REQUIRED,
+                    self.end_headers()
-        tls_version=ssl.PROTOCOL_TLSv1_2,  # Enforce TLS 1.2
+
+        return RequestHandler
+
+
+class TimeoutMonitor:
+    """Background thread to monitor device timeouts"""
+
+    def __init__(self, cache: SensorDataCache, config_module):
+        self.cache = cache
+        self.config = config_module
+        self._should_run = True
+        self.logger = logging.getLogger("TimeoutMonitor")
+
+    def run(self):
+        """Main monitoring loop"""
+        while self._should_run:
+            try:
+                self.cache.check_timeouts()
+
+                # Also cleanup old entries periodically
+                if hasattr(self.config, "cache_cleanup_interval"):
+                    self.cache.cleanup_old_entries(self.config.max_cache_age_hours)
+
+            except Exception as e:
+                self.logger.error(f"Error in timeout monitoring: {e}", exc_info=True)
+
+            # Check every minute
+            time.sleep(60)
+
+    def stop(self):
+        """Stop the monitor"""
+        self._should_run = False
+
+
+def setup_logging(config_module):
+    """Configure logging"""
+    log_level = getattr(logging, config_module.log_level.upper(), logging.INFO)
+    log_format = getattr(
+        config_module,
+        "log_format",
+        "%(asctime)s - %(name)s - %(levelname)s - %(message)s",
+    )
+
+    logging.basicConfig(
+        level=log_level, format=log_format, handlers=[logging.StreamHandler(sys.stdout)]
     )
-    client.tls_insecure_set(False)  # Enforce strict certificate validation
 
-    return client
 
 def main():
-    global mqtt_client
+    """Main application entry point"""
+    # Setup logging
+    setup_logging(config)
+    logger = logging.getLogger("Main")
 
-    # Start timeout monitoring thread
+    logger.info("=" * 60)
-    timeout_thread = threading.Thread(target=monitor_timeout, daemon=True)
+    logger.info("TTN VEGAPULS Air Prometheus Exporter")
-    timeout_thread.start()
+    logger.info("=" * 60)
+    logger.info(f"Integration Method: {config.integration_method}")
+    logger.info(f"Sensor Timeout: {config.sensor_timeout_hours} hours")
+    logger.info(f"HTTP Server: {config.hostName}:{config.serverPort}")
+    logger.info("=" * 60)
 
-    # Start MQTT reconnect thread
+    # Create sensor data cache
-    reconnect_thread = threading.Thread(target=reconnect_mqtt, daemon=True)
+    cache = SensorDataCache(timeout_hours=config.sensor_timeout_hours)
-    reconnect_thread.start()
 
-    while True:
+    # Start timeout monitor
-        mqtt_client = configure_mqtt_client()
+    timeout_monitor = TimeoutMonitor(cache, config)
-        try:
+    monitor_thread = threading.Thread(
-            # Connect to TTN broker
+        target=timeout_monitor.run, daemon=True, name="TimeoutMonitor"
-            broker_url = f"{config.ttn_region.lower()}.cloud.thethings.network"
+    )
-            mqtt_client.connect(broker_url, 8883, 60)
+    monitor_thread.start()
+    logger.info("Started timeout monitor")
 
-            # Subscribe to all topics
+    # Start MQTT client if configured
-            mqtt_client.subscribe("#", 1)
+    mqtt_client = None
-            logging.info(f"Subscribed to all topics.")
+    mqtt_thread = None
+    if config.integration_method.lower() == "mqtt":
+        mqtt_client = TTNMQTTClient(cache, config)
+        mqtt_thread = threading.Thread(
+            target=mqtt_client.run_with_reconnect, daemon=True, name="MQTTClient"
+        )
+        mqtt_thread.start()
+        logger.info("Started MQTT client")
+    else:
+        logger.warning(f"Unsupported integration method: {config.integration_method}")
+        logger.warning("Only 'mqtt' is currently supported")
 
-            poll_mqtt_thread = threading.Thread(target=poll_mqtt, args=((mqtt_client,)))
+    # Start HTTP server
-            poll_mqtt_thread.start()
+    metrics_server = MetricsServer(cache, config)
-        except Exception as e:
+    handler = metrics_server.create_handler()
-            logging.error(f"Error occurred: {e}")
-            mqtt_client.loop_stop()
 
-        webServer = HTTPServer((config.hostName, config.serverPort), RequestHandler)
+    try:
-        print("Server started http://%s:%s" % (config.hostName, config.serverPort))
+        http_server = HTTPServer((config.hostName, config.serverPort), handler)
+        logger.info(
+            f"HTTP server started at http://{config.hostName}:{config.serverPort}"
+        )
+        logger.info("Press Ctrl+C to stop")
 
-        try:
+        http_server.serve_forever()
-            webServer.serve_forever()
+
-        except KeyboardInterrupt:
+    except KeyboardInterrupt:
-            sys.exit(-1)
+        logger.info("\nShutdown requested by user")
+    except Exception as e:
+        logger.error(f"Fatal error: {e}", exc_info=True)
+    finally:
+        # Cleanup
+        logger.info("Shutting down...")
+
+        if mqtt_client:
+            mqtt_client.stop()
+
+        timeout_monitor.stop()
+
+        logger.info("Shutdown complete")
+        sys.exit(0)
 
-            webServer.server_close()
-            print("Server stopped.")
-            poll_mqtt_thread.join()
-        except Exception as e:
-            print(e)
-        time.sleep(60)
 
 if __name__ == "__main__":
     main()

ttn-vegapulsair-exporter.service

@@ -1,16 +1,45 @@
 [Unit]
-Description=TTN Exporter for VEGAPULS Air
+Description=TTN VEGAPULS Air Prometheus Exporter
-After=syslog.target
+Documentation=https://git.mosad.xyz/localhorst/TTN-VEGAPULS-Air-exporter
-After=network.target
+After=network-online.target
+Wants=network-online.target
 
 [Service]
-Restart=on-failure
-RestartSec=2s
 Type=simple
 User=prometheus
 Group=prometheus
+
+# Working directory
 WorkingDirectory=/opt/ttn-vegapulsair-exporter/
+
+# Execution
 ExecStart=/usr/bin/python3 /opt/ttn-vegapulsair-exporter/ttn-vegapulsair-exporter.py
 
+# Restart configuration
+Restart=always
+RestartSec=10
+
+# Logging
+StandardOutput=journal
+StandardError=journal
+SyslogIdentifier=ttn-vegapuls-exporter
+
+# Security settings
+NoNewPrivileges=true
+PrivateTmp=true
+ProtectSystem=strict
+ProtectHome=true
+ReadWritePaths=/opt/ttn-vegapulsair-exporter/
+ProtectKernelTunables=true
+ProtectKernelModules=true
+ProtectControlGroups=true
+
+# Resource limits
+MemoryLimit=256M
+CPUQuota=5%
+
+# Environment
+Environment="PYTHONUNBUFFERED=1"
+
 [Install]
 WantedBy=multi-user.target