Exploring the app

The app is a simple micro-service based on Spring Boot. It provides an API to track the ratings of the talks in real time.

Storage

SQL database with the talks

When the rating is sumbmitted, we must verify that the talk for the given ID is present in our database.

Our database of choice is PostgreSQL 10 accessed with Spring JDBC.

Check com.example.demo.repository.TalksRepository.

Redis

We store the ratings in Redis database with Spring Data Redis (Reactive)

Check com.example.demo.repository.RatingsRepository.

Kafka

We use ES/CQRS to materialize the events into the state. Kafka acts as a broker and we use Spring Kafka.

Check com.example.demo.streams.RatingsListener.

API

The API is a Spring Webflux controller (com.example.demo.api.RatingsController) and exposes two endpoints:

• POST /ratings { "talkId": ?, "value": 1-5 } to add a rating
• GET /ratings?talkId=? to get the histogram of ratings of a talk