- 🟢 Easy to learn — Designed for students and curious minds
- ⚙️ Real-world architecture — Implements RESP and ZeroMQ
- 🎓 Didactic focus — Ideal for teaching databases, systems, and performance fundamentals
This project is a didactic in-memory NoSQL database built using modern C++ (C++20). Developed as part of the C++ MasterClass initiative, it teaches how to implement a real-world key-value store from scratch, covering important systems programming concepts and protocols.
It is fully RESP-compatible, meaning you can use Redis CLI and clients to connect, issue commands like SET, GET, and PING, and receive appropriate responses.
- RESP protocol parsing and serialization — inspired by Redis, implemented in
resp_command_parser.hpp - TCP socket communication via ZeroMQ STREAM sockets — implemented in
server_main.cpp - Pluggable backends (default: STL) for flexible storage — see
backend.hpp,stl_backend.hpp - Extensive command support:
- Strings:
SET,GET,DEL,EXISTS - Sets:
SADD,SREM,SCARD,SMEMBERS,SINTER,SUNION - Sorted Sets:
ZADD,ZSCORE,ZCARD,ZRANGE,ZREM,ZREMRANGEBYSCORE - Database ops:
FLUSHDB,SELECT,DBSIZE,TYPE - Meta:
PING,CLIENT, etc.
- Strings:
- Extensible command execution engine —
execute_command.hpp - ZeroMQ connection monitoring for logging client connections —
zmq_monitor.hpp - Integration-tested using real Redis clients over TCP
| File | Description |
|---|---|
server_main.cpp |
Entry point; sets up ZeroMQ socket, handles RESP traffic |
resp_command_parser.hpp |
Parses RESP messages from Redis clients |
resp_command.hpp |
Holds RESP command abstraction and parsing logic |
execute_command.hpp |
Routes parsed commands to backend strategies |
stl_strategy.hpp |
Implements Redis-like commands using STL containers |
stl_backend.hpp |
Chooses STL as backend and connects context/database/strategy |
backend.hpp |
Backend selector for STL or EASTL |
zmq_monitor.hpp |
Observes client connect/disconnect using ZeroMQ monitor API |
This project serves as a foundation for deeper exploration into key-value store design and distributed systems. Future enhancements may include:
- ➕ Expanding Command Set — Add support for additional Redis-compatible commands to broaden functionality
- 🧩 Plugin Architecture — Introduce a modular plugin system to extend behavior without modifying the core
- 🔐 Authentication Support — Implement
AUTHcommand for client access control and security layers - ⏱️ Key Expiration and TTL — Add time-based key eviction via
TTLandEXPIREsupport - 🚦 Protective Limits — Enforce memory limits, key quotas, and input validation safeguards
- 🧱 Improved Resilience — Enhance stability under high load or malformed requests
- ⚡ Performance Optimizations — Optimize storage and protocol handling for faster throughput
- 🧵 Asynchronous Execution — Support internal async command execution for concurrency and scalability
MIT
Free Software, Hell Yeah!