(Tested on Ubuntu 18.04)
Sensor gateways (including LoRa gateways) and associated online services typically allow the user to define an optional 'HTTP endpoint' or 'HTTP Integration' URL: once this URL is defined, sensor data that arrives at the gateway or online service will be relayed via an HTTP POST of the sensor data (in JSON format) to that URL. This URL must therefore be the address of a server of some sort that is capable of receiving JSON data via HTTP POST and e.g. storing it in a database, graphing it, etc.
Below is a minimal example of such an 'HTTP endpoint' server. The server (here, 'endpoint_server.js') is based on NodeJS; NodeJS is a system that allows for very quick and easy servers to be set up. The server stores and retrieves data to and from an sqlite database; sqlite is a particularly straightforward and lightweight database technology that stores its data in a single file (here, 'db.sqlite').
After installing the necessary prequisites (detailed below), one can run the server (via the command line) directly from the downloaded repository directory by typing:
node endpoint_server.js
This will create a server running which is now listening for HTTP POST requests at the following URL:
[YOUR_IP_ADDRES]:8000/api/endpoint
Incoming POST requests to that URL will be stored it in the sqlite database, if the the JSON data in the POST has the format:
{"temperature":"23.2", "humidity":"10.0", "pressure":"1331"}
The values (23.2,10.0, etc) in the JSON can vary, but the keys (temperature, humidity, etc) cannot.
Note: we provide a helper script, detailed below, that will send a properly-formatted JSON packet via POST to the server.
Any data that is successfully recieved and stored in the database can then later be retrieved by visiting:
[YOUR_IP_ADDRES]:8000/api/latest
-- which will provide the data in the browser in JSON format.
Stored data can also be visualized in a basic graph by visiting:
[YOUR_IP_ADDRES]:8000
More details / functionality detailed below.
This server is based on NodeJS. It's a good idea to use nvm for NodeJS applications, as it allows you to manage multiple versions easily, and also allows for easy un-install if things go awry.
You can download and install nvm (node version manager) with a single line:
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.35.3/install.sh | bash
You may then need to open a new terminal window in order for the 'nvm' command to be available.
Now use nvm to install node version 11:
nvm install 11
nvm use 11
(last line may be redundant).
Sqlite is the database we'll be using to store our data.
To install it, you can use the included script, found inside this repository's 'utilities' folder:
./sqlite3_install.sh
or carry out that script's installation procedure manually:
sudo apt-get install libsqlite3-dev
npm install sqlite3 --build-from-source --sqlite=/usr
npm install
Now we're ready to install the endpoint-server itself. At the top level in this repository, enter the command:
npm install
To run the server, at the top level of this repository, type:
node endpoint_server.js
You will see a message similar to the following:
server running at [YOUR_IP_ADDRESS]:8000
Connected to the SQlite database.
found table 'data'
The most recent entries in the database can be retrieved (in JSON format) via your browser at:
[YOUR_IP_ADDRESS]:8000/api/latest
Graphs (experimental) may be found at:
[YOUR_IP_ADDRESS]:8000
The server is set up to receive JSON-formatted data via an HTTP POST. You can test out this functionality by using the included convenience function, 'post.sh', located in the 'gateway_simulator' folder. This function requires as a command line argument the IP address and port of the running server.
While the server is running in a terminal, open another terminal, and try a test data post:
./post.sh [YOUR_IP_ADDRESS]:8000
Every time that this script is run, it should add another entry to the database.