Skip to content

v4.0.0 - Stable Release

Choose a tag to compare

@Tribrix23 Tribrix23 released this 09 Oct 06:35
· 34 commits to main since this release

CleaveDB v4.0.0 — First Stable Release

CleaveDB is a ground-up, polyglot, hybrid Relational Document & Graph Database designed to eliminate the complexity of traditional SQL JOINs, external vector search services, and separate graph databases.

Instead of juggling three different database technologies for documents, relationships, and vector search, CleaveDB combines all three into a single high-performance engine.


🌟 What Makes CleaveDB Different?

1. Zero-JOIN Relational Graph Bonds

Documents in CleaveDB are not isolated records. They are deeply relational, linked natively through Graph Bonds that allow multi-hop graph traversals in plain, conversational syntax. You can establish bidirectional, expiring, or cascading relationships between documents across different collections without ever writing a complex SQL JOIN.

2. Built From Scratch in Rust (No SQLite, No RocksDB)

The storage core is written entirely in Rust from the ground up. It features custom 16KB fixed-size pages with CRC32 checksums, an in-memory CLOCK-sweep buffer pool cache, a lock-free Write-Ahead Log (WAL) with group commit, and built-in AES-256-GCM encryption at rest.

3. Hardware-Accelerated Vector & SIMD Math

CleaveDB features native C++ AVX-512 and AVX2 vector extensions. Aggregations and mathematical reductions are pushed directly down to the CPU hardware, utilizing 512-bit ZMM registers to process 16 elements per CPU clock cycle, with automatic fallback to standard scalar math on older processors.

4. On-Device AI Semantic Search

Search your documents by meaning, not just exact keyword matches. CleaveDB includes an embedded, quantized ONNX Transformer model (all-MiniLM-L6-v2) that generates 384-dimensional vector embeddings on-device, providing instant semantic relevance search using only ~22MB of RAM.

5. Multi-Layer Distributed Architecture

  • Go Coordinator (coordinator.exe): A standalone microservice that enables high-concurrency Scatter-Gather query orchestration across sharded nodes, merging distributed results using an $O(N \log K)$ Min-Heap algorithm.
  • Rust Edge Gateway (edge_gateway.exe): An asynchronous TCP proxy fronting database connections for high concurrency.
  • Raft Consensus Engine: Built-in multi-node clustering for zero-downtime replication and high availability.

6. CleaveQL Query Language

A case-insensitive, human-readable query language featuring built-in multi-tenancy, field masking, Time-Travel data recovery, and Document-Level Security (DLS).


📦 Getting Started

Download and run CleaveDB-v4.0.0-Setup.exe below to install CleaveDB on Windows 10/11 (64-bit).

  • No prerequisites required: The installer is fully self-contained (~50MB) and runs out-of-the-box without needing Python, Go, Rust, or C++ tools installed on the system.