satasking

Simulation of tasking between a master GroundStation and different Satellite clients.

This project uses a simple backend in Django to serve the GroundStation, Satellite and Task models. Once executed, you can add a GroundStation instance and run it, and as many Satellite instances as you want, and run them.

Once executed the GroundStation instance, it will run a Python SocketServer in background, listening to Satellite (clients) connections. In the other hand, Satellite instances run Python socket clients, and they try to connect to listening SocketServer.

Also you can create as many Task instances as you want (Task takes unique names, a payoff and the resources), and they can be selected and dispatched to be executed by clients. The GroundStation uses a kind of greedy choice algorithm (similar to fractional knapsack problem) to select which Task will be executed and maximize the result payoff.

Here, the satellites are supposed to be multitasking: they can receive several tasks and begin their execution instantly, the only constraint is that received tasks don't compete by resources.

Each satellite "throw a dice" each time a new task arrives to determine if the task will be executed: it raises an error the 10% of time, noticing that the task couldn't be executed.

Once logged in, the API serves the following endpoint: http://localhost:8000/api/taskexecution/ with all detailed information of task dispatching between the groundstation and the client.

Note 1: As this is a very first version, it is mandatory to stop satellite clients before stopping the groundstation server. Otherwise it could hang the django server.

Note 2: There are a lot of TODO commented in the code, they are addressing improval opportunities or technical debt (in some cases).

Installation

Install git and clone this repo. Install Python3 and Pip, then:

  $ pip install -r requirements.txt

For better results, use pyenv and virtualenv.

Backend settings

1. Configure the backend settings by running migrations:

  satasking/ $ python manage.py migrate

2. Create a user:

  satasking/ $ python manage.py createsuperuser

3. Run the tests with:

  satasking/ $ python manage.py test

4. (Optional) Populate DB with fake data with:

  satasking/ $ python load_data.py

Using guide

1. Run the django web server with:

  satasking/ $ python manage.py runserver

This will run the project under http://localhost:8000/

2. Go to http://localhost:8000/admin web interface and create a new GroundStation instance, and many Satellite instances as you want.

In all cases, if they are not specified, by default HOST = 127.0.0.1 and PORT = 65265.

3. Create as many Task instances as you want.
4. Run the created instances, and dispatch the tasks.