The application is hosted on an AWS EC2 instance at http://starkbank-skilltest.info. It consists of a Django application that is served by Nginx as a reverse proxy.
NOTE: In the given time to complete the task, I didn't finish aquiring the needed SSL certificate necessary to secure the site with https. For that reason, the setup to receive the webhook isn't functional. Since the starkbank sandbox API doesn't accept non-encrypted URLs, I would need more time to properly secure the webhook endpoint. I also could not finish writing all the necessary tests.
This application uses the django_cron module to schedule the issuing of 8-12 invoices during a 24h period.
The webhook was created using the starkbank sandbox environment with an invoice subscription. The value of the paid invoices (minus fees) is then transfered to the specified account in the challenge instructions.
As the cronjobs run, detailed logs can be found in app.log
The deployed infrastructure comprises two Docker containers that communicate via a Unix socket. The first container runs a Django-uWSGI worker process responsible for interacting with the Nginx web server to serve dynamic content. These containers communicate seamlessly using shared Docker volumes on the main file system.
Feel free to ask for assistence with the Nginx configuration, as this repository is not so much focused on the server infrastucture, I did not include the files here.
After cloning the repository, set up a python virtual environment and install the dependencies. The python version used was python3.11.2.
pip install -r requirements.txt
The dev_utils.sh script has some useful functions to ease the setup.
Make sure it is executable.
chmod +x dev_utils.sh
And source the contents.
. dev_utils.sh
NOTE: You must configure the host of the EC2 instance in the dev_utils script to properly execute the functions that use ssh to interact remotely.
Ajust the script according to your setup of the file system.
Notice that the redeploy function syncs the production server's code with yours, builds the docker image and remotely executes dkintegration-server.sh.
For this, create the script with the following commands:
docker rm integration-server
docker run -d --restart=unless-stopped \
--name=integration-server \
-v /home/ubuntu/dkdata/integration-server/uwsgi:/uwsgi \
-v /home/ubuntu/dkdata/integration-server/private-key:/private-key \
--env-file=/home/ubuntu/integration-server.env \
-e PRIVATE_KEY=<your starkapi sandbox private key> \
integration-server start.sh
dkdata/integration-server/uwsgi is the path of your unix socket and the application's static files
Use the env file to pass environment variables to the container.
To set up a private key, access https://starkbank.com/sandbox.
ssh on to the hosted instance and run the following script to execute the following script:
docker rm nginx
docker run -d --restart=unless-stopped \
--name=nginx -p 80:80 \
-v /home/ubuntu/dkdata/nginx:/etc/nginx \
-v /home/ubuntu/dkdata/nginx/conf.d:/etc/nginx/conf.d \
-v /home/ubuntu/dkdata/nginx/include:/etc/nginx/include \
-v /home/ubuntu/dkdata/integration-server/uwsgi:/uwsgi_integration-server \
nginx
Notice the common mapping of volumes between the 2 containers.
According to the script, the nginx configuration files for this app should be at ~/dkdata/nginx.