💸 ACID Transactions & Concurrency Simulation in Java

This project demonstrates the ACID properties of database transactions in practice using two approaches:

• Plain JDBC
• Spring JDBC (JdbcTemplate, @Transactional)

Additionally, the project simulates common concurrency issues in databases, such as:

• Dirty Read
• Non-Repeatable Read
• Phantom Read
• Lost Update

---

🏗️ Project Structure

1. BankTransferACIDDemoPlainJDBC

Demonstrates money transfers between accounts using plain JDBC. It fully preserves ACID transaction properties:

• Atomicity: via conn.setAutoCommit(false) and conn.commit()/conn.rollback()
• Consistency: balance validation before transfer
• Isolation: SELECT ... FOR UPDATE locks
• Durability: writes to the transaction_log

---

2. BankTransferACIDDemoSpringJDBC

Implements the same logic, but using Spring:

• @Transactional provides transaction management
• JdbcTemplate simplifies SQL operations

---

3. ParallelExecutionSimulationDemo

Simulates typical issues in multi-threaded environments:

• 🔴 Dirty Read
• 🔁 Non-Repeatable Read
• 👻 Phantom Read
• 💥 Lost Update

---

🧪 How to Run

Requirements:

• Java 17+
• PostgreSQL
• Maven

Database configuration (PostgreSQL):

• Database name: testdb
• Username: admin, Password: admin
• Tables:
  • accounts(account_id, name, balance)
  • transaction_log(account_from, account_to, amount, status)
• Flyway migrates the schema at startup (located in resources/db/migration)

To run:

In Main.java, you can activate specific tests, for example:

BankTransferACIDDemoPlainJDBC.testConcurrentTransfers();
parallelExecutionSimulationDemo.demonstratePhantomRead();

---

📋 Sample Console Output

Transaction 1 Pending  
Transaction 2 Pending  
Transaction 3 Pending  
Transaction 1 completed  
Transaction 2 completed  
Transaction 3 completed  
Account 1: 870.00 PLN  
Account 2: 1130.00 PLN

---

🧠 Conclusions

• ACID properties are crucial in banking and financial systems
• Spring simplifies transaction management, but isolation levels must be chosen consciously
• The simulations show how faulty reads can lead to data inconsistencies

---

📂 Directory Structure

dev.trela
├── ACIDExample
│   ├── BankTransferACIDDemoPlainJDBC.java
│   └── BankTransferACIDDemoSpringJDBC.java
├── parallelexecutionproblems
│   └── ParallelExecutionSimulationDemo.java
├── config
│   ├── AppConfig.java
│   ├── SpringJDBCConfig.java
│   └── PlainJDBCConfig.java
└── Main.java

---

🛠️ Future Enhancements

• Adding a REST API for transfer operations
• Unit testing using H2 or TestContainers
• Transaction and lock monitoring tools