Drone Tracker

An app to report and track location of drones!

Assumptions

• Latitude range: -90.000000 to 90.000000
• Longitude range: -180.000000 to 180.000000
• Drone location data reporting protocol: JSON string sent over UDP
• Test cases cover REST API endpoints and Datagram module (for location reporting)
• Standard ESLint guidelines are used for code syntax
• Data is stored locally in app, under global variable named global.drones
• To prepare for future database IO, async/await and Promise are used in controller and model functions

Reasons for using UDP

UDP was used as protocol of choice because:

• It has a comparatively smaller minimum packet size
• Since location is transmitted every second, Drones can simply transmit their location to server without needing acknowledgement.
• This reduces cellular modem network traffic costs.
• Following this concept, we can further implement other compression methods like gzip to compress location data even more

Container setup

Docker images for the app can be created for two environments: deployment and test. There are two Dockerfiles for each kind of setup in the prjoect directory.

Environment variables

You can optionally set environment vaiables for the following in the Dockerfile:

• HTTP_PORT: Internal port on which HTTP server will run
• UDP_PORT: Internal port on which UDP server will run
• NODE_ENV: This is set to test for unit test mode
• UDP_TIMEOUT: Needed for unit testing of Datagram module, defaults to 30 (value is in milliseconds)

Deployment Mode

Build docker image

docker build -t drone-tracker .

Deploy docker container

The following commands will configure the app to listen to:

• HTTP requests on: http://localhost:8080
• UDP messages on: http://localhost:8081

As a daemon in the background

docker run --rm -p 8080:3000 -p 8081:4000/udp -d drone-tracker

In the foreground

docker run --rm -p 8080:3000 -p 8081:4000/udp drone-tracker

The -p flag is used to bind internal port of container to its external port, where we can actually access the app from outside. The --rm flag removes container after exiting to prevent container and memory clutter. -d denotes whether container is run in daemon mode.

Stop docker container

docker ps
docker stop <CONTAINER ID>

Unit Test Mode

The following Docker commands will build and run the specified npm test script.

docker build -t drone-tracker-test -f Dockerfile.test .
docker run --rm drone-tracker-test

You can also run tests locally with npm run test:local if needed.

Test files are located in /test folder.

API Usage

UDP Datagrams

The application accepts UDP data packets containing drone location information.

The data packet needs to be in stringified JSON format, containing uuid, lat and long values.

e.g.

  {
    "uuid": "my-drone",
    "lat": -31.213123,
    "long": 89.189876
  }

To test the datagram module, you can send UDP messages using the terminal, as follows:

echo '{ "uuid": "my-drone", "lat": -31.213123, "long": 89.189876 }' > /dev/udp/0.0.0.0/8081

REST API

Get a list of:

• All drones: GET: http://localhost:8080/drones
• Inactive drones: GET: http://localhost:8080/drones/incative

You can test these endpoints using curl on the terminal:

curl -G http://localhost:8080/drones
curl -G http://localhost:8080/drones/incative

Size of incoming data packet

When drones report their location over the UDP channel, the size of the message content is logged on the server console in bytes, like so:

  ========== DATAGRAM RECEIVED ==========
    size:   64 bytes
    from:   127.0.0.1:51990
    msg:    { "uuid": "abc", "lat": 31.213123213, "long": -19.18987987988 }
  =======================================

This is obtained from the implementation of the dgram node module. See here for more details. If we need to implement any further allowance logic based on message size, it can be done here.

Lint

• Run ESLint tool: npm run lint
• Autofix some lint issues: npm run lintfix