The GraphRAG Configuration Guide - Overview
22/06/2026
GraphRAG version 1.2.0
This guide covers the Runtime Layer and post-deployment Setup. Before proceeding, complete the base installation and Infrastructure Layer configuration detailed in the official Helm chart README, as they are mandatory prerequisites of the runtime setup.
Important
This guide complements the official Helm chart documentation by focusing strictly on the required configuration steps after the deployment.
The Helm chart README covers the Infrastructure Layer including base installation, image tags, service-to-service URLs, and Kubernetes Secrets for initial connection parameters. The README file is located in charts/graphrag/ folder and contains all relevant instructions.
Tip
This guide only addresses the Runtime Layer and Post-deployment Setup, which determine the GraphRAG workflow's behavior, such as managing of the Workflows configuration, external system endpoints (GraphDB MCP, Graph Modeling), LLM models (primary and fallback), vector store presets, and API credentials within the n8n dashboard. These runtime changes take effect dynamically without requiring a full system or pod redeploy.
This guide focuses on
Service-level configuration for each component of the GraphRAG stack
Connecting GraphRAG to external systems: vector stores, GraphDB, Graph Modeling, Keycloak, and third-party LLM providers
A complete configuration property mapping reference for both deployment structures and the GraphRAG Configuration Workflow
Note
Workflow logic, and error-handling internals are documented separately in the GraphRAG Workflows Guide, including a description of each workflow and its function.