Skip to content

Project.md

Latest commit

 

History

History
112 lines (105 loc) · 4.2 KB

Project.md

File metadata and controls

112 lines (105 loc) · 4.2 KB

Technologies

  • Java 21
  • Spring (Boot, Core, MVC, Security, Data, AOP)
  • Hibernate, Jooq, Liquibase/Flyway
  • PostgreSQL
  • ActiveMQ/RabbitMQ
  • Kafka
  • Hazelcast IMDG
  • Activiti/Camunda BPM
  • JUnit, Mockito, H2, JaCoCo, JBehave
  • Swagger, Open API, REST, gRPC
  • Logback / Log4j2, SL4J
  • JMX, JProfiler, VisualVM
  • Redis, Cassandra
  • Reactive Streams, Akka Actors
  • Docker, Kubernetes, Helm

Project Example

Managing an investment portfolio (example, https://intelinvest.ru), the following functionalities must be implemented:

  • Adding transactions for depositing money, buying or selling securities
  • A dashboard displaying the current value of the portfolio, all transactions, and changes in value
  • The ability to view details about securities
  • An example business process - running every N minutes/hours for each client or for a specific client at the push of a button; steps:
    • Retrieving the value of securities (stocks, bonds) from an external system
    • Recalculating the total portfolio value
    • Recalculating the proportions and returns of individual securities
    • (Optionally) Exporting a brief summary to a file

Project

General Points

  • Using Java 21 and features from versions 8+
  • All modules based on Spring Boot
  • Logging using Logback or Log4j (slf4j)

Module 0

  • Centralized module for authorization/authentication (registration can be added) using Spring Security
  • Creates a user session, stores it in Hazelcast
  • Information about users, roles stored in PostgreSQL
  • Database access through Jooq

Module 1.1

  • Business module, provides REST API
    • Part described in Swagger, generating code and setting up REST
  • Data model - minimum 5-7 related tables
  • Information stored in PostgreSQL - 1 module = 1 database
  • Database scripts deployed using Liquibase
  • Database access through Hibernate and Spring Data JPA

Module 1.2

  • Business module with a process described using BPMN
  • Using Activiti or Camunda
  • Ideally, calling some external system through integration

Module 2.1

  • Collection and aggregation of logs
  • All modules log in Kafka
  • Reading logs and breaking down statistics, for example, the number of messages of each log level per hour
  • Logic implemented using Reactive Streams and Reactor
  • Result written in Redis/MongoDB

Module 2.2

  • Collection and aggregation of logs
  • All modules log in ActiveMQ/RabbitMQ
  • Reading logs and breaking down statistics, for example, the number of messages of each log level per hour
  • Logic implemented using Akka Actor
  • Result written in Cassandra

Module 3

  • Health check verification
  • All modules send their status via gRPC every n seconds (using heap, memory, etc.)
  • Can simply write to log or save in a new database

Frontend

  1. JSP, Thymeleaf, FreeMarker
  2. React, Vue

Functional Testing

  1. Junit + Mockito
  2. Adding H2 + Jbehave

CI/CD

  • Git
  • Using GitHub or GitLab
  • Module 1, distribution - fat jar, app config, log config, start-up script, database update scripts
  • Module 2, distribution - jar, lib, app config, log config, start-up .sh file, database update scripts

Stage 1

  • CI
    • Project build - Maven assembly plugin
    • Run PMD and Checkstyle -> report
    • Run unit tests + coverage percentage with JaCoCo -> report
    • Release version + tag
    • Automated tests (component or integration)
    • Distribution assembly
    • Snapshot version
  • CD
    • Get familiar with Docker
    • Set up Jenkins in Docker - pull project from Git, build, add test, Docker image, store in Docker registry
    • Configure a job that takes the distribution (ZIP), creates folders/updates version
      • Install distribution - ZIP with jar, configs, etc.
      • Update relational database using Liquibase
      • Consider new deployment (folder structure) and update (logs preserved)
      • Restart/launch service

Stage 2

  • CI
    • Assemble distribution -> Create Docker Image
  • CD
    • Deploy artifacts -> Launch container(s) using docker-compose (consider updating individual modules)

Stage 3

  • CI
    • Create Docker Image -> Create Helm Chart
  • CD
    • Launch containers -> Deploy to Kubernetes

Load Testing

  • Write a script in bash/python

to generate load on the business module

  • Analyze behavior of all modules and database using jconsole, visualvm, jprofiler