steam-stack

A simple school projet to mimic steam - Docker compose stack

• steam-stack
  • Docker profiles
  • How to run
    • How to run the steam stack itself
    • How to run steam stack related apps
  • Details
  • Known issues

Docker profiles

This docker compose project has the following profiles:

• postgres (for all databases in this project)
• kafka (to initialize the cluster)
• ksqldb (to initialize ksqldb for the cluster)
• publishers (for all publisher related images)
• platform (for all platform related images)
• players (for all players related images)
• all (to start/stop everything)

How to run

How to run the steam stack itself

First initialize the postgres databases:

docker compose --profile postgres up -d

Then just initialize the Kafka cluster:

docker compose --profile kafka up -d

Now you have everything needed to run the other images that would listen to the published topics ! Check details for ksqldb specific instructions.

To stop everything, just use:

docker compose --profile all down

How to run steam stack related apps

The following apps:

• player-app
• pub-app
• platform-app

can be connected to this stack. Use the following command and adjust if needed:

docker run --rm -it --network kafka-net -e SPRING_DATASOURCE_URL=jdbc:postgresql://pub-db:5432/pub --name pub-app ghcr.io/mortell560/pub-app:latest

Check the relevant repositories for more info...

Take note that the platform-app is already provided to you through the all profile. You can check the platform service declared within the docker-compose.yml

Important: take note that you can edit every topic and group-id that an app uses through environment variables or Spring Boot configurations with keys such as CLI_GROUP or TOPIC_USER_TO_REGISTER... Edit if needed.

Details

In this docker compose stack, you will find a few folders:

• scripts (for all schema registry and kafka-connect actions)
• schemas (which contain all AVRO schemas used in this project)
• kafka-connect/plugins (contains all the Apache Kafka Connectors used in this project)
• ksql (for all "SQL" scripts intended for KSQLDB)

Important note: ksqldb may be a bit picky and may not start with the entire cluster. You can start it later on with:

docker compose --profile ksqldb up -d

If you still don't see any stream and table within Kafka UI declared for KSQLDB, try again to initialize it:

docker compose --profile ksqldb up -d ksqldb-init

Known issues

• Please start Ksqldb ONLY when you have already published a game and you are about to buy a game. There's a mismatch in the schemas that ksqldb tries to declare and the Kafka Connectors AVRO schema's (public.games.first-games-price topic). There are no issues if you launch KSQLDB after the schema has been declared already by Kafka Connect.