Here is the general schematic for our system's architecture.
Firstly, We need to go cd kafka/mongodb-kafka-base then open a terminal and run docker-compose -p mongo-kafka up.
This will creates needed mongodb kafka setup for us.
Then, We need to enable sink connector for log collection from kafka to mongodb. Open a terminal then run docker exec -it mongo1 /bin/bash.
We need to copy sink connector from kafka/mongodb-kafka-base/simplesink.json via use nano like nano simplesink.json and save the file.
Then, We can configure with existing tool on image cx simplesink.json. Our configuration is done.
Secondly, We need to run a rabbitmq image, open a terminal and run docker run -it --rm --name rabbitmq -p 5672:5672 -p 15672:15672 -e RABBITMQ_DEFAULT_USER=admin -e RABBITMQ_DEFAULT_PASS=123456 rabbitmq:3.13-management
We need mysql for our application. For simple testing, we can use single mysql container. But in production, we can seperate database containers and kafka log collection.
Open a terminal and run docker run -d --name some-mysql -p 3306:3306 -p -e MYSQL_ROOT_PASSWORD=password mysql.
We need to configure database for one time. Open another terminal and run docker exec -it some-mysql /bin/bash.
Enter mysql -uroot -p -A command and enter password.
Then we need to run commands at initial_db.sql.
Initial setup is complete. We can start all services and use via gateway localhost:8084 or directly to localhost:8080.
Each bank has own service, including creating new loan and listing the loans endpoints. Also each bank has own application repository.
This service can create user, list user and has user repository. But when user needs to check their applications or create new application, connects to each bank that registered on service, fetchs applications or creates applications over client.
We can collect all logs via Kafka then saved to the MongoDB.
We have a basic notification service that consumes the queue on the RabbitMQ.
