NebulaGraph is an open-source distributed graph database built for super large-scale graphs with milliseconds of latency. It delivers high performance, scalability, and availability for storing and processing graph data.

This Helm chart deploys a NebulaGraph cluster on Kubernetes.

The NebulaGraph Operator must be installed first — see the nebulagraph-operator directory.

Installation

To install NebulaGraph, run:

make install

Usage

After installation:

# Check if NebulaGraph cluster is running
kubectl get pods -n nebula

# Access Graphd service
kubectl port-forward svc/nebula-graphd -n nebula 9669:9669

Configuration

The default configuration includes:

NebulaGraph cluster with 3 graphd, 3 metad, and 3 storaged nodes
Persistent storage for data
Default port configurations

Components

Graphd: Query engine for graph processing
Metad: Metadata management
Storaged: Storage engine for graph data

Connecting to NebulaGraph

Use Nebula Console to connect:

# Install nebula-console
# Connect to the cluster
./nebula-console -addr 127.0.0.1 -port 9669 -u root -p nebula

Graph Data Model

NebulaGraph uses:

Vertices: Represent entities
Edges: Represent relationships
Tags: Define vertex types
Edge Types: Define relationship types

Uninstall

To uninstall NebulaGraph:

make uninstall