The purpose of this application is to showcase a mechanism that tracks
user UI interactions on a payment page.
So imagine a situation where you have a percentage of users
whose transactions are being canceled/rejected within your provided payment page.
We're interested in finding out what was happening before those transactions were canceled/rejected
while a user was surfing that particular payment page.
| Frontend | Backend | Database |
|---|---|---|
| React JS | Spring Boot | MongoDB |
- We have two types of connections:
- Traditional Restful connection.
- Tracker events are sent from the UI to BE via a websocket connection.
- This websocket connection is intended to be separate & dedicated & always open for each payment window.
- The backend side including websocket connection is implemented reactively via
spring:webflux. - Each time a user loads the payment page, there'll be a payment session assigned to it.
- Payment sessions are created with id of the
websocket's session id(managed by spring framework). - Exposing tracking data via a
RESTAPI. MongoDBis our choice as a data-source for its nice archiving capabilities.