Monitor your devices using metrics and see beautiful dashboards in grafana

TLDR; Prometheus is an open-source monitoring system with a dimensional data model, flexible query language, efficient time series database and modern alerting approach.

Alert manager Alertmanager handles alerts sent by the Prometheus server.

Grafana Grafana is the open source analytics & monitoring solution for every database

Combine these tools with Offbeat-IoT to monitor your devices and see beautiful dashboards using docker

Observe your device through prometheus metrics

Get a better overview of the sensors on your devices by polling metrics and viewing graphs in grafana.

If a sensors measurements is out of limits, then have the alert manager send you an email, chat or call a webhook.

Lets imagine you have a device that measures the temperature and humidity in a room.

If it sends that data to [Offbeat-Iot](https://offbeat-iot.com) over websockets, like so:

{
  "temperature"=18,
  "humidity":80
 }

Then when you call this endpoint and use the username/password of a websocket user

https://offbeat-iot.com/api/devices/<the endpoint id of your device>/prometheus

Then you will see data like this:

# HELP temperature 
# TYPE temperature gauge
temperature{device="friendly name of the device",} 18.0
# HELP humidity
# TYPE humidity gauge
humidity{device="alexa-switch",} 80.0
# HELP connected Shows if device is connected. 1.0 = true/0.0 = false
# TYPE connected gauge
connected{device="friendly name of the device",} 1.0
# HELP connected_since_total 
# TYPE connected_since_total counter
connected_since_total{device="friendly name of the device",} 0.0

We see that two [gauges](https://prometheus.io/docs/concepts/metric_types/#gauge) are created for the measurements.

We also see that two other metrics are shown: * connected * This also a gauge, which shows if the device is connected or not * connected_since * This is a [counter](https://prometheus.io/docs/concepts/metric_types/#counter) which shows how long the device has been connected with millisecond precision.

Assuming you have [git](https://git-scm.com/) and [docker](https://www.docker.com/) installed, you can configure prometheus by cloning this repository and running the compose file

git clone git@github.com:Offbeat-IoT/monitoring-example.git

and then by creating some volumes to store data in the stack

docker volume create offbeat-iot-prometheus-data
docker volume create offbeat-iot-grafana-data
docker volume create offbeat-iot-alertmanager-data

then open prometheus/data/offbeat-iot-devices.yml and add your websocket users username and password at the bottom of the file.

When all this is done, you are ready to start the stack.

docker compose up -d

The -d parameter starts the stack as a deamon and you have your hands free in the console

If you open your browser and navigate to

http://localhost:9090/

you have entered your prometheus instance

and if you then head to status→targets you can see the targets prometheus is polling

If they are not all a delightful blue color, then they probably will be after a few browser refreshes.

If you navigate to the Graph and start typing temperature or humidity then you should see the data collected from Offbeat-Iot, from your device.

If you then navigate to

http://localhost:3000

You will enter grafana. Where you can log in using admin/admin as user and password. Grafana will as you to set a more secure password.

Then, you can update the example dashboard with your own device.

Or, [create your own dashboard](docs/new-dashboard.md)