How I Built a Resilient Idempotency Engine for Financial Transactions (Vanilla Java)

How I Built a Thread-Safe Idempotency Engine for High-Resilience Systems Building robust backend systems means preparing for the worst-case scenario. One of the most common issues in financial or critical systems is the Duplicate Request problem. A user clicks a button twice, a network retry triggers a second call, and suddenly, you have a double-spending issue. To solve this, I developed a lightweight Idempotency Engine in Java, focusing on a "Vanilla-First" approach to maintain total control over memory and performance. ⚙️ The Architecture The engine acts as a State Interceptor . It sits between the incoming request and the business logic. If a request with the same unique key arrives, the engine knows exactly how to handle it based on its previous state. Key Features: LRU Cache (Least Recently Used): I used a synchronized cache to manage transaction records in memory. This prevents Memory Overflow by discarding the oldest entries under heavy load. Thread-Safety: By using Collections