README.md

BME280 Data Logger with faasd

This data logger writes measurements from a Bosch BME280 or BMP280 sensor into a InfluxDB time-series database.

You'll need one Raspberry Pi 3 or 4 to run the data collection stack and dashboard, then as many additional Raspberry Pi Zeros as you like for each location you want to monitor.

The hardware

How to connect the sensor over i2c

For the hardware, follow the instructions from the growlab app under "Assembling the build". You'll need the same parts, but you can skip the camera module.

Deployment

You will run faasd on your Raspberry Pi 2, 3 or 4 to store data readings and to run OpenFaaS and Grafana.

On your Raspberry Pi Zero, or whichever host has a sensor connected to it, you'll run the sender app.

Deploy faasd

Deploy faasd to your Raspberry Pi 3 or 4 using these instructions

Customise the password, and update /var/lib/faasd/docker-compose.yaml

  influxdb:
    image: docker.io/library/influxdb:1.8
    environment:
      - INFLUXDB_DB=defaultdb
      - INFLUXDB_ADMIN_USER=admin
      - "INFLUXDB_ADMIN_PASSWORD=PASSWORD"
      - INFLUXDB_USER=user
      - "INFLUXDB_USER_PASSWORD=PASSWORD"
      - INFLUXDB_REPORTING_DISABLED=true
      - INFLUXDB_HTTP_AUTH_ENABLED=true
      - INFLUXDB_HTTP_BIND_ADDRESS=0.0.0.0:8086
    volumes:
      # we assume cwd == /var/lib/faasd
      - type: bind
        source: ./influxdb/
        target: /var/lib/influxdb
    user: "1000" 
    cap_add:
      - CAP_NET_RAW
    ports:
      - "0.0.0.0:8086:8086"

Make a directory for InfluxDB:

mkdir -p /var/lib/faasd/influxdb
chown 1000:1000 /var/lib/faasd/influxdb

Then reload and restart:

sudo systemctl daemon-reload \
  && sudo systemctl restart faasd

Deploy the function

Create a secret for the InfluxDB user:

export PASSWORD=""

faas-cli secret create influx-pass --from-literal $PASSWORD
faas-cli secret create influx-user --from-literal admin

Set the influx_db env-var in stack.yml i.e. 192.168.0.21

Next, deploy the function:

faas-cli deploy

Build if you like:

faas-cli publish -f stack.yml --platforms linux/arm/7

Deploy the sender onto your Raspberry Pi with a sensor

On your Raspberry Pi with the sensor, you'll run the "sender" app.

Enable i2c and change the hostname as required using the raspi-config tool.

If you haven't installed the main growlab app for live-previews, then install the below dependencies:

sudo apt update -qy && \
  sudo apt install -qy python3 \
  i2c-tools \
  python3-pip \
  git \
  tmux

Clone the growlab app:

git clone https://github.com/alexellis/growlab
cd growlab/data-logger/sender/

Install any pip modules required for the sender app:

pip3 install -r requirements.txt

Then run the sender app:

FUNCTION_URL=http://192.168.0.21:8080/function/submit-sample \
  SENSOR=my-shed \
  python3 main.py

Note: if you're using a BMP280 sensor then add an addition environment variable of SENSOR=bmp280

A sensor reading will be submitted to faasd every 30 seconds. You can alter this sample interval by editing sender/main.py.

Going further with a dashboard

Deploy Grafana to faasd using the instructions in the eBook Serverless For Everyone Else

Then create yourself a simple dashboard for the measurements you see in the "readings" database.

Once you have it up and running, create a datasource, then import the dashboard.json file and open the dashboard to view your sensor readings.

My very cold shed - measured overnight!

A larger version in the summer time: