RadishDB

RadishDB is a Redis-inspired, in-memory key–value database written in C, built to understand how real databases work internally and how to deploy them like production services.

RadishDB implements real storage engine internals and is fully containerized and deployable using Docker and Docker Compose.

This project focuses on:

• write-ahead logging (WAL)
• crash recovery
• expiration & TTL
• log compaction
• persistent storage
• containerized deployment (Docker)
• service orchestration (docker-compose)

RadishDB is a learning-grade but architecturally serious database, following patterns used in Redis, RocksDB, and PostgreSQL.

---

Screenshots

Interactive REPL

TCP Server Mode

---

Core Features

Key–Value Storage

• String → string mapping
• Hash table with separate chaining
• Automatic resizing
• Heap-managed memory
• Safe insert, update, delete

---

TTL & Expiration

• Per-key expiration timestamps
• Passive expiration on read
• Active expiration via sweeper
• Redis-style TTL semantics

---

Persistence

RadishDB uses two persistence layers.

Snapshot (.rdbx)

Binary snapshot of full database state.

Used for fast save and restore.

---

Append-Only File (AOF)

Crash-safe write-ahead log.

Features:

• deterministic replay
• fsync durability
• partial write protection
• crash recovery

---

AOF Rewrite (Log Compaction)

RadishDB implements log compaction:

• serializes current state
• writes new compact log
• removes obsolete history

This ensures:

• bounded disk usage
• fast startup
• crash safety

---

Architecture

RadishDB is built as a protocol-agnostic execution engine.

engine.c      → command semantics
hashtable.c   → storage engine
aof.c         → durability
expires.c     → TTL system
persistence.c → snapshots
repl.c        → interactive mode
server.c      → TCP server
main.c        → lifecycle

This separation allows multiple frontends:

• REPL
• TCP server
• container deployment
• future protocols

---

Supported Commands

SET key value
SET key value EX seconds
GET key
DEL key
TTL key
COUNT
SAVE file
LOAD file
BENCH n
INFO
HELP
EXIT

---

Example Session

SET name radish
OK

GET name
radish

SET color red EX 10
OK

TTL color
7

After restart:

GET name
radish

State recovered from AOF.

---

Build and Run (Native)

Build:

make

REPL mode:

./radishdb

Server mode:

./radishdb --server

Server listens on port 6379.

---

Run Using Docker (Recommended)

RadishDB is published on Docker Hub.

Pull image:

docker pull piee314/radishdb:latest

Run container:

docker run -d \
  --name radishdb \
  -p 6379:6379 \
  -v radish-data:/app/aof \
  piee314/radishdb:latest

This starts RadishDB with:

• persistent storage
• exposed network port
• isolated container runtime

---

Persistence

RadishDB stores data in:

/app/aof/radish.aof

Docker volume ensures data persists across container restarts.

List volumes:

docker volume ls

Inspect volume:

docker volume inspect radish-data

---

Run Using Docker Compose

docker-compose.yml:

version: "3.8"

services:
  radishdb:
    image: piee314/radishdb:latest
    container_name: radishdb
    ports:
      - "6379:6379"
    volumes:
      - radish-data:/app/aof
    restart: unless-stopped

volumes:
  radish-data:

Start service:

docker-compose up -d

Stop service:

docker-compose down

View logs:

docker-compose logs

---

Container Architecture

RadishDB uses multi-stage Docker build:

build stage → compile binary
runtime stage → minimal Linux container
volume mount → persistent storage

Benefits:

• small image size
• fast startup
• production-ready deployment

---

DevOps Features Demonstrated

RadishDB demonstrates real DevOps practices:

• Docker containerization
• multi-stage builds
• persistent volumes
• Docker Hub image publishing
• docker-compose orchestration
• reproducible deployments

RadishDB can run identically on:

• local machine
• cloud servers
• CI/CD pipelines
• container orchestration platforms

---

Project Structure

src/
├── engine.c
├── hashtable.c
├── aof.c
├── expires.c
├── persistence.c
├── repl.c
├── server.c
├── result.c
├── utils.c

Dockerfile
docker-compose.yml
Makefile
README.md
ARCHITECTURE.md
LEARNINGS.md

---

Deployment Example (Cloud Server)

On any Linux server:

docker run -d \
  -p 6379:6379 \
  -v radish-data:/app/aof \
  piee314/radishdb:latest

RadishDB runs remotely and persists data.

---

Status

Stable learning release.

Fully functional storage engine with persistent container deployment.

---

License

MIT License

---

Author

Built as a systems engineering and DevOps learning project to understand:

• database internals
• crash recovery
• persistence
• containerization
• deployment workflows