FlintDB
Projects — — Links: Github Repo — Documentation
FlintDB is a self-engineered, flat-file NoSQL database designed to master core backend principles. This project focuses on implementing data atomicity and complex architectural features to prove a deep understanding of reliable data transaction logic.
The Problem
Most lightweight flat-file data solutions fail when a program crashes during a write operation, corrupting the data. FlintDB was designed as a capstone to solve this crucial integrity issue by engineering transactional support that ensures data is either fully written or not written at all, preventing data corruption.
Key Contributions
- Developed custom transaction logic to enforce Atomicity (the ‘A’ in ACID), ensuring that multi-step write operations are treated as a single, indivisible unit to prevent data corruption.
- Engineered a Transparent Data Encryption (TDE) system and implemented a Two-Tier Key System to secure data at rest, showcasing crucial security principles.
- Built a custom, file-based caching system to manage resource access and optimize read performance, demonstrating efficiency in systems design.
- Designed the entire data structure and file management system, showcasing proficiency in building complex logical components from the ground up.
Technology Used
- Python, PHP: The core languages for the entire system.
- JSON: The underlying data storage format.
