tradik/mddb

By tradik

•Updated 9 days ago

High-performance markdown database with gRPC/HTTP/3 - 29,810 docs/sec, 5.75x faster than MongoDB

Image

10K+

Overview Tags

tradik/mddb repository overview

⁠MDDB - High-Performance Markdown Database

The fastest markdown database with gRPC and HTTP/3 support

MDDB is a specialized, high-performance database server designed for storing and managing markdown documents with rich metadata, full revision history, and dual protocol support (HTTP/JSON + gRPC/Protobuf).

⁠🚀 Performance

29,810 docs/sec throughput (37.4x baseline)
34µs average latency
5.75x faster than MongoDB
6.89x faster than PostgreSQL
24.54x faster than MySQL
95.43x faster than CouchDB

⁠⚡ Key Features

Dual Protocol Support: HTTP/JSON REST API + gRPC/Protobuf
HTTP/3 Support: QUIC protocol for extreme performance
Full Revision History: Track all document changes with MVCC
Rich Metadata: Store and query structured metadata
Batch Operations: High-throughput batch insert/update/delete
Compression: Snappy and Zstd compression support
Lock-Free Cache: Sharded cache with zero contention
Adaptive Indexing: Smart metadata indexing
Built-in Backup: Hot backup and restore functionality
Embedded Database: BoltDB for zero-dependency deployment

⁠🐳 Quick Start

⁠Production Server

docker run -d \
  --name mddb \
  -p 11023:11023 \
  -p 11024:11024 \
  -v mddb-data:/data \
  -e MDDB_PATH=/data/mddb.db \
  -e MDDB_MODE=wr \
  -e MDDB_EXTREME=true \
  tradik/mddb:latest

⁠Docker Compose

version: '3.8'

services:
  mddb:
    image: tradik/mddb:latest
    container_name: mddb-server
    ports:
      - "11023:11023"  # HTTP
      - "11024:11024"  # gRPC
      - "11443:11443"  # HTTP/3 (extreme mode)
    volumes:
      - mddb-data:/data
    environment:
      - MDDB_PATH=/data/mddb.db
      - MDDB_MODE=wr
      - MDDB_EXTREME=true
      - MDDB_HTTP_PORT=11023
      - MDDB_GRPC_PORT=11024
      - MDDB_HTTP3_PORT=11443
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:11023/stats"]
      interval: 30s
      timeout: 10s
      retries: 3

volumes:
  mddb-data:

⁠🔧 Configuration

⁠Environment Variables

Variable	Default	Description
`MDDB_PATH`	`/data/mddb.db`	Database file path
`MDDB_MODE`	`wr`	Mode: `ro` (read-only) or `wr` (read-write)
`MDDB_EXTREME`	`false`	Enable extreme performance mode
`MDDB_HTTP_PORT`	`11023`	HTTP server port
`MDDB_GRPC_PORT`	`11024`	gRPC server port
`MDDB_HTTP3_PORT`	`11443`	HTTP/3 server port (extreme mode)

⁠Extreme Performance Mode

Enable with MDDB_EXTREME=true to activate:

HTTP/3 server with QUIC protocol
All 29 performance optimizations
Lock-free cache with sharding
Adaptive indexing
Compression (Snappy + Zstd)
Optimized batch processing
String allocation elimination
Zero-copy operations

⁠📊 Available Tags

⁠Production Images

latest - Latest stable release
2.0.1, 2.0, 2 - Specific version tags

⁠Platform Support

All images support:

linux/amd64 (Intel/AMD x86_64)
linux/arm64 (ARM/Apple Silicon aarch64)

⁠🔌 API Examples

⁠HTTP REST API

# Add a document
curl -X POST http://localhost:11023/add \
  -H "Content-Type: application/json" \
  -d '{
    "collection": "docs",
    "key": "welcome",
    "lang": "en",
    "meta": {
      "title": {"values": ["Welcome"]},
      "author": {"values": ["MDDB Team"]}
    },
    "content_md": "# Welcome to MDDB\n\nHigh-performance markdown database."
  }'

# Get a document
curl http://localhost:11023/get/docs/welcome/en

# Search documents
curl -X POST http://localhost:11023/search \
  -H "Content-Type: application/json" \
  -d '{
    "collection": "docs",
    "filter_meta": {
      "author": {"values": ["MDDB Team"]}
    },
    "limit": 10
  }'

# Batch insert (high performance)
curl -X POST http://localhost:11023/batch \
  -H "Content-Type: application/json" \
  -d '{
    "collection": "docs",
    "documents": [
      {
        "key": "doc1",
        "lang": "en",
        "meta": {"title": {"values": ["Document 1"]}},
        "content_md": "# Document 1"
      },
      {
        "key": "doc2",
        "lang": "en",
        "meta": {"title": {"values": ["Document 2"]}},
        "content_md": "# Document 2"
      }
    ]
  }'

# Get statistics
curl http://localhost:11023/stats

⁠gRPC API

import (
    "google.golang.org/grpc"
    "mddb/proto"
)

conn, _ := grpc.Dial("localhost:11024", grpc.WithInsecure())
client := proto.NewMDDBClient(conn)

// Add document
doc, _ := client.Add(ctx, &proto.AddRequest{
    Collection: "docs",
    Key:        "welcome",
    Lang:       "en",
    Meta: map[string]*proto.MetaValues{
        "title": {Values: []string{"Welcome"}},
    },
    ContentMd: "# Welcome to MDDB",
})

// Batch insert (high performance)
resp, _ := client.AddBatch(ctx, &proto.AddBatchRequest{
    Collection: "docs",
    Documents: []*proto.BatchDocument{
        {
            Key:  "doc1",
            Lang: "en",
            Meta: map[string]*proto.MetaValues{
                "title": {Values: []string{"Document 1"}},
            },
            ContentMd: "# Document 1",
        },
        // ... more documents
    },
})

⁠🏗️ Architecture

┌─────────────────────────────────────────────────────┐
│                 Performance Layer                    │
│  • HTTP/3 Server (QUIC)                             │
│  • Lock-Free Cache (Sharded)                        │
│  • Batch Processor (Parallel + Single TX)           │
│  • String Optimization (Zero-Copy)                  │
│  • Compression (Snappy + Zstd)                      │
└─────────────────────────────────────────────────────┘
                         ↓
┌─────────────────────────────────────────────────────┐
│                  Protocol Layer                      │
│  • HTTP/JSON REST API                               │
│  • gRPC/Protobuf API                                │
│  • HTTP/3 (Extreme Mode)                            │
└─────────────────────────────────────────────────────┘
                         ↓
┌─────────────────────────────────────────────────────┐
│                   Storage Layer                      │
│  • BoltDB (Embedded)                                │
│  • MVCC (Multi-Version Concurrency Control)         │
│  • Adaptive Indexing                                │
│  • Revision History                                 │
└─────────────────────────────────────────────────────┘

⁠📈 Benchmarks

Tested with 3,000 documents on identical hardware:

Database	Throughput	Avg Latency	vs MDDB
MDDB (Batch)	25,863 docs/s	39µs	1.00x
MongoDB	5,452 docs/s	182µs	4.74x slower
PostgreSQL	3,360 docs/s	297µs	7.69x slower
MySQL	1,183 docs/s	845µs	21.86x slower
CouchDB	310 docs/s	3,201µs	83.39x slower

⁠🔐 Security

No authentication by default (add reverse proxy for production)
Runs as non-root user in container
Read-only mode available (MDDB_MODE=ro)
Volume permissions: 750 (owner: mddbd)

⁠📦 Volume Management

# Create named volume
docker volume create mddb-data

# Backup database
docker run --rm \
  -v mddb-data:/data \
  -v $(pwd)/backups:/backups \
  tradik/mddb:latest \
  cp /data/mddb.db /backups/mddb-backup-$(date +%Y%m%d).db

# Restore database
docker run --rm \
  -v mddb-data:/data \
  -v $(pwd)/backups:/backups \
  tradik/mddb:latest \
  cp /backups/mddb-backup-20250107.db /data/mddb.db

⁠🛠️ Troubleshooting

⁠Check logs

docker logs mddb

⁠Check health

curl http://localhost:11023/stats

⁠Performance issues

Enable extreme mode: MDDB_EXTREME=true
Use batch operations for bulk inserts
Use gRPC for better performance than HTTP
Increase Docker resources (CPU/Memory)

⁠Database locked

Ensure only one instance is running
Check file permissions on volume
Use read-only mode for multiple readers

⁠📚 Documentation

GitHub: https://github.com/tradik/mddb⁠
API Documentation: https://github.com/tradik/mddb/blob/main/README.md⁠
Examples: https://github.com/tradik/mddb/tree/main/test⁠
Changelog: https://github.com/tradik/mddb/blob/main/CHANGELOG.md⁠

⁠🤝 Support

Issues: https://github.com/tradik/mddb/issues⁠
Discussions: https://github.com/tradik/mddb/discussions⁠
Email: [email protected]⁠

⁠📄 License

MIT License - see LICENSE⁠

⁠🌟 Why MDDB?

Performance First: 29 optimizations for maximum throughput
Developer Friendly: Simple API, easy deployment
Production Ready: Battle-tested, stable, reliable
Zero Dependencies: Embedded database, no external services
Modern Protocols: HTTP/3, gRPC, Protobuf
Rich Features: Metadata, revisions, search, backup
Open Source: MIT licensed, community driven

Made with ❤️ by the MDDB Team

Star us on GitHub⁠ if you find MDDB useful!

Tag summary

Recent tags

Content type

Image

Digest

sha256:8fdc3fbe2…

Size

35.5 MB

Last updated

9 days ago

docker pull tradik/mddb