sreehari-aot / IOT-server

This project aims to store the data collected from an IOT device into a time series database, QuestDB for example and create graphical representations of the data to better understand the data.

IOT-server

This project aims to store the data collected from an IOT device into a time series database, QuestDB for example and create graphical representations of the data to better understand the data.

A small chunk of datasets are provided by the server which the publisher will use to push.

The publisher script act as an IOT device which publishes data points in certain intervals.

This project consistes of 4 majour components

1. Database:- The database used is Quest DB which is a time series database. This is where we publish our data from IOT devices. The database is running on port 9000 To make sure database is up and running go to http://localhost:9000

2. Server:- The server is a nodejs application which exposes endpoints for publishing and retrieving data from the database. The server is running on port 8000 To check if the server is up and running go to http://localhost:8000 after starting the application. Make sure the database is up and running before starting server. (ignore if using compose file)

   Server endpoints

   POST
   url path: {server-url}/publish eg: localhost:8000/publish
   body: {
       "timeStamp":"1662896469284",
       "person": "1",
       "pos_x": 15.32,
       "pos_y": 8.75,
       "vel_x": 0,
       "vel_y": 0,
       "o_id": "631dad359fbc895818809423"
   }
   
   payload type: application/json
   
   GET
   
   url path: {server-url}/get
   query params: 
     query {url encoded SQL query}
     type enum [1 / 2] 1- for position graph, 2 - for count graph
   
   eg: http://localhost:8000/get?query=SELECT%20*%20from%20%27Human%27&timings=true&type=1
   
   
   

3. Publisher:- The publisher is a script which mocks an IOT device by publishing observations everu 20 seconds. Make sure the server is up and running before running publisher script. Once the script is started leave it running since it will publish the data over time and exit once all data is published.

4. Client:- The client is a react application which helps to visualize the data in a graphical format. The client is running in port 3000 To visit the client go to http://localhost:3000

   The client offers different charts

   • Line charts
   • Heatmaps (TODO)

   The line chart supports 2 options - position (default) - count The default selection in linechart representst pos_x over different timestamps The count selection represents the number of persons at a purticular time

Requirements

• Docker
• Nodejs: v14.17.0

Get started

Using Docker compose

1. run docker-compose up -d to spin up the databasee, server, client and the publisher

The publisher will start sending data once the container is up and running, be patient and wait for couple of minutes to see the graphical representation of the data, head over to http://localhost:3000 to get the charts and will be updated with new data point every 20 secs. Please refresh to see the updated data.

for development

1. clone the repo
2. run docker-compose up -d to start the database
3. Before next steps make sure to stop client and server containers since they will use the allocated porst preventing dev server to run.
4. check into server directory
5. run npm i && npm start
6. once the server is up and running
7. open up a new terminal and check into publisher directory
8. run npm i && node script.js and leave it running
9. check into client directory
10. run docker-compose up -d / npm i && npm start to start the client

TODO:

• Heatmap chart for the data points
• Move hardcoded values into configurations
• Single compose file to setup entire application
• Optimized script to publish large data sets
• Script modification to add time delay to mock an IOT device

Dependencies

• react:18.2.0
• recharts:2.3.2
• react-heatmap-grid:0.9.0
• express:4.18.2
• @questdb/nodejs-client:1.0.2