Skip to content

Latest commit

 

History

History
224 lines (163 loc) · 7.45 KB

Home.md

File metadata and controls

224 lines (163 loc) · 7.45 KB

Welcome to ThemisDB

A high-performance multi-model database with ACID guarantees

Version CI License

🚀 NEW in v1.4.0-alpha (January 5, 2026)

Advanced LLM Features:

  • 📝 Grammar-Constrained Generation - EBNF/GBNF support for guaranteed valid outputs (95-99% reliability)
  • 🔭 RoPE Scaling - Extended context window from 4K → 32K tokens (8x increase)
  • 🖼️ Vision Support - Multi-modal LLMs with CLIP-based image encoding
  • ⚡ Flash Attention - CUDA kernels for 15-25% speedup, 30% memory reduction
  • 🎯 Speculative Decoding - 2-3x faster inference with draft+target models
  • 🔄 Continuous Batching - 2x+ throughput with dynamic request batching

Enterprise Features:

  • 🔥 Hot Spare Management - Automatic failover with health monitoring
  • 📊 Enhanced Prometheus Metrics - LLM monitoring and cache performance
  • 🔄 WAL Replication via gRPC - Distributed inter-shard replication
  • 🎮 Multi-GPU LoRA Support - Distributed LoRA adapters
  • 🐘 PostgreSQL Protocol - COPY, prepared statements, transactions

Documentation:

Overview

ThemisDB is a production-ready multi-model database that combines relational, graph, vector, and document models in a single system with full ACID transaction support. Built on RocksDB with advanced security and compliance features.

Key Capabilities:

  • 🔒 ACID Transactions - Full snapshot isolation with MVCC
  • 🔍 Multi-Model - One database for relational, graph, vector, and documents
  • 🚀 High Performance - 45K writes/s, 120K reads/s
  • 🛡️ Enterprise Security - TLS 1.3, RBAC, encryption, audit logging
  • 🌐 Distributed - Horizontal sharding, replication, Kubernetes-ready
  • 🧠 AI-Ready - Hybrid search, embedding cache, GPU-accelerated

Quick Links

🚀 Getting Started

📖 Learn ThemisDB

🚀 Deploy to Production

Core Features

Multi-Model Database

Relational:

  • Secondary indexes (equality, composite, range)
  • SQL-like AQL queries
  • ACID transactions

Graph:

  • Native graph storage
  • BFS, Dijkstra, A* traversals
  • Path constraints and pruning

Vector:

  • HNSW and FAISS indexes
  • GPU-accelerated similarity search
  • Hybrid search for RAG workflows

Document:

  • JSON storage with flexible schema
  • Fast field extraction
  • Schema-based encryption

Advanced Analytics

  • CEP Engine - Complex Event Processing with pattern matching
  • OLAP - CUBE, ROLLUP, window functions
  • Time-Series - Gorilla compression, continuous aggregates
  • Streaming - Real-time data processing

Enterprise Security

  • Authentication - RBAC with 4-tier hierarchy, mTLS
  • Encryption - AES-256-GCM at rest, TLS 1.3 in transit
  • Audit - 65+ event types, SIEM integration
  • Compliance - GDPR, SOC 2, HIPAA ready
  • Secrets - HashiCorp Vault integration

Distributed Capabilities

  • Sharding - Consistent hashing, 150 virtual nodes
  • Replication - Leader-follower and multi-master
  • Redundancy - RAID-like modes (MIRROR, STRIPE, PARITY)
  • Kubernetes - Operator with CRDs
  • Monitoring - 44 Prometheus metrics, Grafana dashboards

Performance Benchmarks

Operation Throughput Latency (p50)
Entity PUT 45,000 ops/s 0.02 ms
Entity GET 120,000 ops/s 0.008 ms
Indexed Query 8,500 queries/s 0.12 ms
Graph Traverse 3,200 ops/s 0.31 ms
Vector ANN (k=10) 1,800 queries/s 0.55 ms

Full Benchmarks →

Documentation Structure

This wiki is organized into the following sections:

For Users

  • Getting Started - Installation, quick start, configuration
  • Features - Detailed feature documentation
  • Query Language - AQL syntax and examples
  • API Reference - REST, GraphQL, client SDKs

For Operators

  • Operations - Deployment, monitoring, backup
  • Security - TLS, RBAC, encryption, compliance
  • Performance - Tuning and optimization

For Developers

  • Development - Building, testing, contributing
  • Architecture - System design and internals
  • Advanced Topics - Sharding, GPU, plugins

Roadmap

Completed (v1.0 - v1.2)

  • ✅ ACID transactions with MVCC
  • ✅ Multi-model support (all 4 models)
  • ✅ Horizontal sharding and replication
  • ✅ GPU acceleration (10 backends)
  • ✅ Enterprise security features
  • ✅ Client SDKs (7 languages)
  • ✅ Hypertables and hybrid search

Current Focus (v1.3.0 - Q1 2026)

  • 🚧 Query optimizer v2
  • 🚧 RE2 integration for security
  • 🚧 SDK publishing (PyPI, npm, crates.io)
  • 🚧 Penetration testing phase 1

Planned (v1.4.0+ - 2026)

  • 📋 Multi-datacenter deployment
  • 📋 Advanced ML/GNN features
  • 📋 DuckDB OLAP integration
  • 📋 Real-time materialized views

Full Roadmap →

Community & Support

Quick Start Example

# Pull and run with Docker
docker pull themisdb/themisdb:latest
docker run -d -p 8765:8765 themisdb/themisdb:latest

# Create an entity
curl -X PUT http://localhost:8765/entities/users:alice \
  -H "Content-Type: application/json" \
  -d '{"blob":"{\"name\":\"Alice\",\"age\":30}"}'

# Query
curl -X POST http://localhost:8765/query \
  -H "Content-Type: application/json" \
  -d '{"table":"users","predicates":[{"column":"age","value":"30"}]}'

Full Quick Start →

License

ThemisDB is open source under the MIT License.

Ready to get started? → Quick Start Guide

Need help? → Discussions