Tesladata

This toolset polls data from the Tesla API and puts it in a Mongo database.

The main components of Tesladata are the poller (poller.py) and the watcher (watcher.py).

Poller

The poller component reads data from the Tesla API and stores it a Mongo database.

Watcher

The watcher component connects to the Mongo database where the Poller stored the data and when a record is inserted, 'watcher' sends it to InfluxDB and publishes it on a MQTT topic.

The data is stored in InfluxDB so that it can be plotted with Grafana and published on MQTT for integration with Home-Assistant.

Configuration

Edit the file config.yaml, the entries should be self-explanatory.

Installation

Clone this repo and install teslajon from https://github.com/gglockner/teslajson and follow the 'Installation' part. Then run pip install -r requirements.txt in the tesladata directory to install the rest of the Python requirements.

Run make install to install Tesladata on your system and start the poller service with the command systemctl start tesladata-poller.service

MongoDB

Running a MongoDB can easily be done as a Docker container. To run a test database (the data will be lost when the container is stopped):

docker run --rm --name mongodb -p 27017:27017 -d mongo --replSet rs0 --oplogSize 1024

Or if you want to run it and store the data permantly, add a directory as a volume (this is how I run it):

docker run --name mongodb -p 27017:27017 -v /my/own/datadir:/data/db -d mongo --replSet rs0 --oplogSize 1024

MongoDB replica set

When the MongoDB is started for the first time, you need to initiate a replica set. Do this with the command rs.initiate() in the mongo shell. To get a mongo shell on your docker container, run sudo docker exec -ti mongodb mongo

This replica set is needed, even if you have one MongoDB server, the Mongo function db.watch() we use, needs the MongoDB 'oplog' that is created/enabled when the recordset is initiated.

Tools

mongo-to-influxdb.py

Read data from MongoDB and put it in InfluxDB. Use this to (re)fill InfluxDB with missing datapoints.

Usage:

./mongo-to-influxdb.py --configfile </path/to/config.yaml> --secondsback <integer>

The secondsback argument is used to generate a timestamp in milliseconds which is used to fetch documents from the MongoDB that are newer than that timestamp.

mongo-dump.py

Read data from the MongoDB and return it as JSON on stdout

Usage

usage: mongo-dump.py [-h] [--mongoserver MONGOSERVER] [--database DATABASE]
                    [--collection COLLECTION] [--diff]

optional arguments:
  --mongoserver     the mongoserver address [localhost]
  --database        the mongo database name where the Tesla data is [tesla]
  --collection      the mongo collection you want to see [all]
  --diff            show difference between stored data points [False]

sleepy-to-sleeping.py

Measures the time between state 'sleepy' and 'asleep' and inserts the results in InfluxDB. Run this once per hour, missing datapoints can be inserted by running mongo-to-influxdb.py