Laporte

• a gateway for processing metrics with automation and states
• acceptor for sensors, batch jobs or messages
• with the ability to set actuators
• compatible with Prometheus for metric alerting, collecting and data visualization

Schema:

Features:

• can evaluate, recalculate or correct metric
• calculate new metric or state from other metrics
• automation using a true algorithmization + data structures (python expression code)
• can set TTL for metrics (obsolete metric disappears when time is over)
• communicates via:
  • RESTful API
  • realtime, bidirectionally using Socket.IO
  • exporting data as Prometheus metrics
  • live www status page
• optional bridges (extensions):
  • laporte-mqtt to connect a large family of devices using MQTT protocol
  • laporte-journal to store all events into other database

Installation:

To install Laporte from PyPI:

pip install laporte

Or using container image from Docker Hub:

docker pull vinklat/laporte

Quick HOWTO:

Example: push gateway for weather sensor

• manually simulate a weather sensor and send temperature and relative humidity data to this gateway using REST API
• in this example the temperature will be corrected (slope +1% and shift + 0.5). Can be used in case of calibrate sensor, unit conversion, ...
• other two new metrics will be calculated (dew point temperature and absolute humidity)
• in the case of an inactive sensor, the data expires in 10 minutes

a) run the docker image

docker run -p 9128:9128 vinklat/laporte -c conf/example_weatherstation.yml

(the content of this built-in config file can be seen here: example_weatherstation.yml)

b) send metrics from sensor via REST API

Try sending data several times using curl:

curl http://localhost:9128/api/metrics/weather1 -d temp_celsius=37.5 -d hum_ratio=0.8 -X PUT

c) watch status

• Laporte status page: http://localhost:9128
• JSON response of REST API: http://localhost:9128/api/metrics/by_node
• Prometheus metrics: http://localhost:9128/prom

...more info on the wiki