DingoDB is a distributed, multi-modal vector database designed for high concurrency and low-latency queries on structured and unstructured data. It targets developers and enterprises seeking a unified platform that integrates relational semantics with vector search, offering MySQL compatibility and built-in high availability.

How It Works

DingoDB combines traditional relational database indexing with various vector index types for scalar-vector hybrid retrieval. It utilizes a distributed storage system (Dingo-Store) based on the Raft protocol for strong consistency. The architecture supports automatic elastic data sharding, real-time index optimization, and tiered retrieval for massive datasets, including disk-based vector search to minimize memory usage.

Quick Start & Requirements

• Install: Docker or Ansible deployment options are available.
• Prerequisites: Java Profiler tools like YourKit are recommended for performance-critical development.
• Documentation: All Documentation, Install, Usage, Developing

Highlighted Details

• Unified SQL (MySQL-Compatible) interface for structured and unstructured data.
• First-class citizen data models for Table and Vector.
• Built-in, external-component-free high availability and automatic elastic data sharding.
• Real-time scalar and vector index building with background optimization.

Maintenance & Community

DingoDB is sponsored by DataCanvas. Community feedback is welcomed. Contact information is provided for support and suggestions.

Licensing & Compatibility

Licensed under the Apache License Version 2.0. This license is permissive and generally compatible with commercial and closed-source applications.

Limitations & Caveats

The project is actively developed, with a focus on core features. Specific performance benchmarks or detailed scalability limits are not explicitly stated in the README. Development recommendations include using VS Code and YourKit Java Profiler.