Introduction

This project aims to provide an API for retrieving account movements, applying concepts like async communications, observability and use of container orchestration. It is built upon the foundation of the existing repository here and is designed to run seamlessly in a Kubernetes environment.

Getting Started

Project to expose, through an API, the report of movements from the accounts. Transactions are created by the Seed console application and published to a topic in RabbitMQ, then read by the Movements.AsyncReceiver application, which processes and saves the data in the PostgreSQL database. This data is then exposed through the Movements.Api at the /report/{{accountId}} endpoint.

Setup in docker-desktop

1 - execute script run.sh to create all required resources

# use parameter --build-push to build and push docker images to docker hub
./run.sh # --build-push

2 - Access movements public API

• Docs movements API - Swagger

  http://localhost/movements/swagger

• Get movements report for account 123456-78

  curl http://localhost/movements/v1/report/123456-78

3 - Access logs in kibana

• Local Kibana address

  http://localhost/kibana

• Configuring index pattern for see logs:
  1. Index patterns page
  2. Create data view
  3. Name: fluentd-logs
  4. Create data view button

4 - Clean up, run script clean-up.sh

./clean-up.sh

Version running in AWS EKS

transactions-aws-eks

Architecture

• transactions-seed: Responsible for reading fake data about transactions from a JSON file and then publishing data to a topic
• movements-async-receiver: Responsible for reading events from a queue attached to the topic and saving these movements in a Postgres SQL database
• movements-api: REST API responsible for exposing reports of account movements for an account
• nginx ingress: Responsible for redirecting traffic to the right locations
• fluentd: Responsible for tailing log files and sending them to Elasticsearch
• kibana: Interface responsible for querying and visualizing logs in Elasticsearch

Technologies

• C# was used as the language with .net 6, following some of the concepts of clean architecture. For unit tests, xunit and moq were used.
• Docker was used for the application containers with kubernetes for container orchestration.
• PostgreSQL was chosen as the database.
• RabbitMQ was chosen as the message broker.
• Fluentd was used for log aggregation, sending the logs to Elastic Search.
• Kibana was used for log visualization.
• GitHub Actions were used for CI while the application was being developed, built, and tested on each push.
• Kubernetes was used to deploy and host all components