Skip to content

Latest commit

 

History

History
114 lines (82 loc) · 3.83 KB

DEPLOYMENT.md

File metadata and controls

114 lines (82 loc) · 3.83 KB
Raw

A) Fully local setup

To setup and test service locally clone repository and make by command make

After make is complete activate virtual environment: source .venv/.../activate

Before you can run the service, you need to do the following:

  1. Create a local Postgres database
  2. Edit webserver/settings.py, add the following:
DATABASES = {
    'default': {
        'ENGINE': 'django.db.backends.postgresql_psycopg2',
        'NAME': '<db_name>',
        'USER': '<username>',
        'PASSWORD': '<password>',
        'HOST': 'localhost',
        'PORT': '5432',
    }
}

  1. Set the REDIS_URL environment variable, or add the following line to init_django.py: os.environ.setdefault("REDIS_URL", 'redis://localhost:6379/0') (if running Redis locally)

  2. Set the SECRET_KEY environment variable or change the default to non-empty string in webserver/settings.py

  3. Make and run migrations (assuming you have activated the virtual env):

python manage.py makemigrations
python manage.py migrate

  1. Run webserver/create_user.py. This will give you an API key that you will use for API queries.

  2. Make sure Redis server is running locally (type redis-server in the terminal to start it).

Once these steps are complete, run the following two processes:

python manage.py runserver
celery worker -app=tasks.app

You can now query the API using your Binance keys and the user API key returned by create_user.py.

B) Creating heroku app

If user is not authenticated on heroku use heroku login.
To create heroku app use heroku create command. If everything goes right, you will see Creating app... done, ⬢ young-above-487 where young-above-487 is the application name.

Adding databases

In heroku the databases are used as additions. To see if there are databases attached simply run heroku addons.

PostgreSQL

By default heroku creates postgres instance and the URI is stored in heroku environment as DATABASE_URL. If it is not present then we need to add it manually. heroku addons:create heroku-postgresql:[plan] -a [application_name] For development purposes it is better to pick hobby-dev plan which is free. For example heroku addons:create heroku-postgresql:hobby-dev -a black-unicorn-123

Redis

For celery workers we need to add redis manually. Again by using addons command heroku addons:create heroku-redis:[plan] -a [application_name]

After these steps we have created heroku app, added postgres and redis.

Deployment

Deploying on Heroku

To deploy service remotely just run git push heroku master. And now it is ready! You can open app in browser by heroku open.

In order to scale the web or worker instances just heroku ps:scale web=1 or heroku ps:scale worker=1 correspondingly.

To turn off the instance put the value 0 heroku ps:scale worker=0.

Deploying locally (optional)

Local deployment is useful for debugging.

To deploy service locally we need to export postgres and redis URIs to our env. For getting URIs just run heroku run env | grep DATABASE and heroku run env | grep REDIS and export returned values export [returned_value]. Then just run by command heroku local. Congratulations you have running instance on your computer.

Initial set up

Setting up the database for a fresh install of postgreSQL

If you are using a fresh install, you'll need to initialize the database

heroku run python manage.py migrate

Creating new users

Creating user can be done by webserver/create_user.py script. To create user call heroku run python webserver/create_user.py

It will return the api_key of created user. You will see a promt with the new api key

% heroku run python webserver/create_user.py
Running python webserver/create_user.py on ⬢ black-unicorn-123... up, run.2577 (Free)
Created a user with "77777aaaaaaaaaaaaaaaaaeaaaa77777" API key.