Simple meteor application for monitoring temperature and humidity in a room. Gathering the measurements themselves is done using a separate Python script, see Gathering data.

You only need a server with node.js, npm and mongodb for deploying this application. Refer to this article for a rather detailed tutorial on how to do this.

I run this application on a Raspberry Pi Zero with Arch Linux. You could create a simple start script, such as the one below, and create a systemd service that runs this script for you. The script is based on the README file inside the bundle directory:

#!/bin/bash

export MONGO_URL='mongodb://localhost:27017/<database>'
export ROOT_URL='http://<as-desired>'
export PORT=80

cd /your/path/to/bundle
node main.js

Simply run meteor build . on your development machine and upload the resulting tar-ball to your server. Expand the tar-ball there, and run the following script (for example):

#!/bin/bash

BASE_DIR=/your/home/directory/
DEFAULT_USER=your-username
SERVICE_NAME=climate-monitor

if [[ $UID != 0 ]]; then
	echo "This deployment script requires sudo rights for systemctl."
	exit 1
fi

# Stop the server
systemctl stop ${SERVICE_NAME}

# Remove existing files
rm -rf ${BASE_DIR}bundle/

# Extract the new package (as default user)
sudo -u ${DEFAULT_USER} tar xfz room-climate-monitor.tar.gz -C ${BASE_DIR}

# In case you have any npm build issues, you may want to comment out the following lines (dirty fix):
## Fix package conflicts (we want fibers 1.0.8 for build purposes)
#cd ${BASE_DIR}bundle/programs/server
#sed -i 's/1\.0\.5/1.0.8/g' npm-shrinkwrap.json
#sed -i 's/\"fibers\"\: \"1\.0\.5\"/"fibers": "1.0.8"/g' package.json

# Install npm packages (as default user)
sudo -u ${DEFAULT_USER} npm install

# Start the server again
systemctl start ${SERVICE_NAME}

Just booting this system won't give you any data, it is merely a (simple) visualisation of a MongoDB collection. I use an AM2302 sensor wired to the GPIO pins of the Pi, together with the Adafruit DHT library to insert measurements directly into the MongoDB collection (using pymongo).

The data format used is straight-forward, namely a document with three fields: ts containing the timestamp (as integer, in milliseconds), temperature containing the measured temperature (as float, or integer as desired) and humidity containing the measured humidity (as float, or integer as desired).