The storm alert microservice for the R nineT Scrambler navigation tower implements the Secret Law of Storms set of rules which can be found on worldstormcentral.co.
Data is collected from the MQTT topics weather/pressure and weather/temperature which is published by the
Weather Station Service.
Whenever there is an alert condition, the service will publish it to the MQTT topic alarm/weather which may
then be handled by other services. Please see the source code for the detailed format of the alert message.
The service will send the last 24 hours of barograph data (4 data points/hour - 15 minute intervals) to the
MQTT topic weather/barograph when requested.
To request barograph data send the message { fetch: true } to the topic storm/barograph.
This service provides the following restful endpoints:
/storm/barograph- returns the last 24 hours of barometric readings (4 data points/hour)
Persistent data storage for logging atmospheric pressure and temperature is done on a SQLite database and the Node.JS module sqlite3 is used to access it.
The current version of sqlite3 must be installed from the source on the Raspberry PI Zero and will take a very long time to complete. To install the database engine and Node.JS module run:
sudo apt-get install sqlite3 libsqlite3-dev
npm install sqlite3 --build-from-source --sqlite=/usr/local
After installation, the SQLite database schema should be created by running the following command:
sqlite3 weather.db < weather.sql
To run the service on system startup, the PM2 process manager for Node.JS applications is used. It provides process monitoring capabilities which are great for this application. To start the service run:
pm2 start app.js --name storm --watch --time --ignore-watch="weather.*" -- --port=3000