Skip to content

Latest commit

 

History

History
73 lines (56 loc) · 3.18 KB

README.md

File metadata and controls

73 lines (56 loc) · 3.18 KB
Raw

maracker

API aiming to make Human Brain Project's Medical Informatics Platform's developped apps deployment on Mesos Marathon easier.

Build Status

Requirements

  • PostgreSQL
  • Python 3.6+

The docker images you want to enter into the database must be indexed on the MicroBadger platform as Maracker depends on this API.

Here's a guide explaining how to keep your container's metadatas up-to-date in MicroBadger using a webhook.

Installation for local testing

Assuming you already have a working instance of PostgreSQL and Python 3.6 or later installed and your operating system is GNU/Linux, you can install the API by doing the following actions:

  1. python -m venv venv to create a virtual environment.
  2. source venv/bin/activate to activate the environement.
  3. pip install -r requirements.txt to install the project's dependencies.
  4. Create a database named maracker in PostgreSQL or choose another name.
  5. If you choose another database name and your database user is not postgres, set the following environment variables using export so that Django will have the right database connection information:
    • DB_NAME
    • DB_USER
    • DB_PASSWORD
    • DB_HOST
    • DB_PORT
  6. cd maracker && python manage.py migrate to create the database.
  7. python manage.py runserver to run the API.
  8. You can test the API and query it with curl http://localhost:8000/mipapps which lists the available MIP applications in the database.
  9. If you want to add a new application, you can test it using the test.json file which adds the MIP's woken container's data in the database. Just run curl -X POST http://localhost:8000/mipapps -d @test.json -H "Content-Type: application/json".

If you want to run the unit tests, use the manage.py script and run python manage.py test.

Execution details

The infrastructure (Mesos, Marathon, Chronos and Traefik) can be launched by executing the run.sh script. This launches docker containers and restarts the traefik container 50 seconds after the others have been started so that it connects to Marathon correctly. This amount of time can be set by modifying the run.sh script.

The following UIs are accessible after the script executed:

You can then try to deploy applications on Marathon using either the examples in tests/deployable-marathon-apps or yours.

If you want to test the developped API, you can try to send the JSON files in tests/maracker-use-cases (docker-nginx.json, docker-webapp.json or docker-webapp2.json) to the API.

If you you want to stop the stack, launch docker-compose -f stack-compose.yml stop.