Back to Nodes

Mode

Last updated May 20, 2026

n8n community node for openGauss DataVec vector database

49 Weekly Downloads
870 Monthly Downloads

Included Nodes

Mode
Resource

Description

n8n-nodes-opengauss-datavec

An n8n community node package for openGauss DataVec — provides both an AI Vector Store node for RAG/agent workflows and a general-purpose SQL node for standard database operations.

Features

Vector Store Node (openGauss DataVec Store)

AI Vector Store sub-node that integrates with n8n AI Agent, Retriever, and other LangChain nodes.

  • Get Many — Vector similarity search, returns Top-K results
  • Insert Documents — Insert documents with embeddings (auto-creates table on first run)
  • Retrieve Documents (As Vector Store) — Use as a Vector Store Retriever sub-node
  • Retrieve Documents (As Tool for AI Agent) — Use as an AI Agent tool node
  • Configurable: Schema, Table, Dimensions, Distance Strategy, Top K, Metadata Filter.

    SQL Node (openGauss)

    General-purpose execution node for standard database operations.

  • Execute SQL — Run arbitrary SQL queries
  • Insert / Update / Upsert / Delete / Select — CRUD operations with a visual UI
  • Configurable: Schema, Table, Columns, WHERE conditions, etc.

    Supported Distance Metrics

    | Metric | Description |
    |——–|————-|
    | L2 | Euclidean distance |
    | Cosine | Cosine similarity |
    | Inner Product | Dot product similarity |
    | Manhattan | Manhattan (L1) distance |

    Supported Index Types

    | Index | Description |
    |——-|————-|
    | HNSW | Hierarchical Navigable Small World graph index |
    | IVFFLAT | Inverted file with flat quantization |
    | DISKANN | Disk-based approximate nearest neighbor |

    Prerequisites

  • openGauss 7.0+ with DataVec extension enabled
  • n8n instance (self-hosted or cloud)

Installation

Via n8n Community Nodes

1. Go to Settings > Community Nodes in your n8n instance
2. Search for n8n-nodes-opengauss-datavec
3. Click Install

Via npm

npm install n8n-nodes-opengauss-datavec

Quick Start

See MIT