OctoPrint-Prometheus-Exporter

This is a utility plugin, which enables the prometheus server to scrape metrics from your octoprint instance. Later on, you can use data vizualisation tools (for example grafana) to track and visualize your printer(s) status(es).

This plugin has no visible UI!

Currently exported metrics:

• python version - as info
• octoprint version, hostname, os - as info
• actual temperature - as gauge with tool identifier label
• target temperature - as gauge with tool identifier label
• client number - as gauge; the actually connected clients to the host
• printer state - as info
• started prints - as counter
• failed prints - as counter
• done prints - as counter
• cancelled prints - as counter
• timelaps count - as counter
• print progress - as gauge with path label
• slice progress - as gauge with path label
• print total time - as counter
• last print time - as gauge
• fan speed - as gauge
• extrusion total - as counter
• x, y and z travel - as a counter
• last print extrusion - as gauge
• print time elapsed - as gauge
• print time estimate - as gauge
• print time left estimation - as gauge

All of the metrics are prefixed as octoprint_ for easier identification.

The metrics endpoint is: http://localhost:5000/plugin/prometheus_exporter/metrics (change the host+port to your actual host+port)

Setup

Install via the bundled Plugin Manager or manually using this URL:

https://github.com/tg44/OctoPrint-Prometheus-Exporter/archive/master.zip

Prometheus config

Add this to the scrape_configs part of your prometheus.yml:

- job_name: 'octoprint'
    scrape_interval: 5s
    metrics_path: '/plugin/prometheus_exporter/metrics'
    static_configs:
      - targets: ['octoprint:80']

Or if you have enabled authentication:

 - job_name: 'octoprint'
    scrape_interval: 5s
    metrics_path: '/plugin/prometheus_exporter/metrics'
    params:
      apikey: ['__OCTOPRINT_APIKEY__']
    static_configs:
      - targets: ['octoprint:80']

Permission system

New in version 0.2.0: by default all users / operators have access to the metrics endpoint. If you want to make metrics accessible to anonymous users (guests) without disabling your entire authentication system simply add the metrics permission to the guest user group.

Local development/testing

dev.sh

Run bash.dev in this repository. This will

• Create a Python virtual environment venv/
• Install Octoprint OctoPrint>=1.4.0
• Install this plugin
• Configure Octoprint
  • Create a user. admin:admin
  • Skip all wizards
  • Enable virtual printer and autoconnect
• Start OctoPrint in debug mode.

Docker

There is a docker-compose file, which will start:

• an octoprint instance on port 5000
  • !!! you should install the plugin to it
  • TODO: auto load the plugin to the newly created container
  • !!! you should add virtual printer to it (after the first start docker/octoprint_data/config.yaml and restart the container)
  • without docker:
    • clone
    • cd to the dir
    • virtualenv --python=/usr/bin/python3 venv3
    • source venv3/bin/activate
    • pip install "OctoPrint>=1.4.0"
    • pip install -e .
    • octoprint serve --debug
    • (on mac there could be a hidden "give me a root password" line)
• a prometheus instance on port 9090
  • configured to pull the metrics from the octoprint
• a grafana instance on port 3000
  • configured prometheus as datasource
  • user/pass is admin/foobar

You can start the stack with a regular docker-compose up -d.