Ragtech Supervise Docker Image with Prometheus metrics

Ragtech is a Brazilian company that produces UPS (more commonly known in Brazil as no-break) devices. They have a software called Supervise that is used to monitor and control the UPS devices.

The existing Supervise software packaging is not really suitable with modern Linux distributions. This project aims to provide a Docker image with Supervise supplying all required dependencies.

Also, a Prometheus metrics endpoint was added to easily export metrics data.

Usage

You can run a new container from the computer where the UPS USB cable is plugged to. Validate that the serial interface created by this USB is named /dev/ttyACM0 and replace it accordingly:

$ docker run -d --name supervise --device /dev/ttyACM0:rw -p 4470:4470 -p 4471:4471 ghcr.io/valtlfelipe/ragtech-supervise:latest

Docker Compose

services:
  ragtech-supervise:
    image: ghcr.io/valtlfelipe/ragtech-supervise:latest
    devices:
      - /dev/ttyACM0:/dev/ttyACM0:rw
    ports:
      - "4470:4470"  # Supervise web interface and API
      - "4471:4471"  # Prometheus metrics exporter
    volumes:
      - ./data:/data

Logging

All log output is written to stdout and stderr. Logs are categorized with 5 different prefixes:

• init: Logs related to the container initialization/termination
• main: Logs referring to the stderr/stdout of the main supsvc process
• supsvc: Logs related to the main supervise functionality
• device-manager: Unknown, but I assume it's related to the communication with the UPS'es
• serialhid: Logs related to the Serial interface to the UPS

Interface

The container exposes two HTTP ports:

• Port 4470 — Ragtech Supervise web interface and API
• Port 4471 — Prometheus metrics exporter (this container)

Web Interface

Access the web interface at http://localhost:4470 in your browser.

Prometheus Metrics

A Prometheus-compatible metrics endpoint is available at http://localhost:4471/metrics.

Available metrics:

Metric	Type	Description
ragtech_ups_status	Gauge	UPS connection status (1=connected, 0=disconnected)
ragtech_ups_input_voltage_volts	Gauge	Input voltage in volts
ragtech_ups_output_voltage_volts	Gauge	Output voltage in volts
ragtech_ups_output_current_amps	Gauge	Output current in amps
ragtech_ups_output_frequency_hertz	Gauge	Output frequency in Hz
ragtech_ups_output_power_watts	Gauge	Output power in watts
ragtech_ups_battery_charge_percent	Gauge	Battery charge percentage
ragtech_ups_battery_voltage_volts	Gauge	Battery voltage in volts
ragtech_ups_temperature_celsius	Gauge	UPS temperature in Celsius
ragtech_ups_load_percent	Gauge	Load as percentage of nominal power
ragtech_ups_led_red	Gauge	Red LED state (0 or 255)
ragtech_ups_led_green	Gauge	Green LED state (0 or 255)
ragtech_ups_led_blue	Gauge	Blue LED state (0 or 255)
ragtech_system_uptime_milliseconds	Gauge	System uptime in milliseconds since epoch
ragtech_collector_scrape_duration_seconds	Gauge	Duration of the last scrape
ragtech_collector_scrape_errors_total	Counter	Total number of scrape errors

All UPS metrics include device_id and device_name labels.

Health check:

$ curl http://localhost:4471/health
OK

Example Prometheus scrape config:

scrape_configs:
  - job_name: 'ragtech-supervise'
    static_configs:
      - targets: ['localhost:4471']

SQLite Database

Alternatively, have programatic access to the UPS data by querying the underlying SQLite database.

IMPORTANT: The SQLite database is set to use WAL as the journaling mode, so you can read the database while it's being written to. Because of that, you need to also account for all the database files:

• /data/monit.db
• /data/monit.db-wal
• /data/monit.db-shm

This is how you would run the container with the database mounted to the host filesystem:

$ mkdir host-db-path
$ docker run -d --name supervise --device /dev/ttyACM0:rw -p 4470:4470 -p 4471:4471 -v ./host-db-path:/data ghcr.io/valtlfelipe/ragtech-supervise:latest

License

Apache 2.0