LightRAG/docs/archives/action_plan/multitenant-audit/01-test-protocol.md
Raphael MANSUY 2b292d4924
docs: Enterprise Edition & Multi-tenancy attribution (#5)
* Remove outdated documentation files: Quick Start Guide, Apache AGE Analysis, and Scratchpad.

* Add multi-tenant testing strategy and ADR index documentation

- Introduced ADR 008 detailing the multi-tenant testing strategy for the ./starter environment, covering compatibility and multi-tenant modes, testing scenarios, and implementation details.
- Created a comprehensive ADR index (README.md) summarizing all architecture decision records related to the multi-tenant implementation, including purpose, key sections, and reading paths for different roles.

* feat(docs): Add comprehensive multi-tenancy guide and README for LightRAG Enterprise

- Introduced `0008-multi-tenancy.md` detailing multi-tenancy architecture, key concepts, roles, permissions, configuration, and API endpoints.
- Created `README.md` as the main documentation index, outlining features, quick start, system overview, and deployment options.
- Documented the LightRAG architecture, storage backends, LLM integrations, and query modes.
- Established a task log (`2025-01-21-lightrag-documentation-log.md`) summarizing documentation creation actions, decisions, and insights.
2025-12-04 18:09:15 +08:00

8.5 KiB

Multi-Tenant Test Protocol

Date: November 29, 2025


Test Environment Setup

Prerequisites

  1. macOS System Requirements

    • Python 3.12+ installed
    • Node.js 18+ installed
    • Docker Desktop installed and running
  2. Repository Setup

    cd /Users/raphaelmansuy/Github/03-working/LightRAG
    git checkout feat/multi-tenannt
    

Step 1: Start Database Services (Docker)

Create a docker-compose file specifically for testing databases:

# docker-compose.test-db.yml
services:
  postgres:
    image: pgvector/pgvector:pg16
    container_name: lightrag-test-postgres
    environment:
      POSTGRES_USER: lightrag
      POSTGRES_PASSWORD: lightrag123
      POSTGRES_DB: lightrag_test
    ports:
      - "5432:5432"
    volumes:
      - postgres_test_data:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U lightrag -d lightrag_test"]
      interval: 5s
      timeout: 5s
      retries: 5

  redis:
    image: redis:7-alpine
    container_name: lightrag-test-redis
    ports:
      - "6379:6379"
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 5s
      timeout: 5s
      retries: 5

volumes:
  postgres_test_data:

Start databases:

docker-compose -f docker-compose.test-db.yml up -d

Step 2: Configure Environment

Create .env.test file:

# Database configuration
POSTGRES_HOST=localhost
POSTGRES_PORT=5432
POSTGRES_USER=lightrag
POSTGRES_PASSWORD=lightrag123
POSTGRES_DATABASE=lightrag_test

# Redis configuration
REDIS_HOST=localhost
REDIS_PORT=6379

# Multi-tenant configuration
MULTITENANT_MODE=demo
INIT_DEMO_TENANTS=true

# LightRAG configuration
LIGHTRAG_WORKING_DIR=./data/rag_storage
LIGHTRAG_INPUT_DIR=./data/inputs

# LLM Configuration (for testing)
LLM_PROVIDER=openai
LLM_MODEL=gpt-4o-mini
OPENAI_API_KEY=your-key-here

# Embedding Configuration
EMBEDDING_PROVIDER=openai
EMBEDDING_MODEL=text-embedding-3-small
EMBEDDING_DIM=1536

# API Configuration
LIGHTRAG_API_KEY=test-api-key-123
AUTH_USER=admin
AUTH_PASS=admin123

Step 3: Initialize Python Environment

# Create virtual environment
python -m venv .venv-test
source .venv-test/bin/activate

# Install dependencies
pip install -e ".[dev,postgres,redis]"

Step 4: Start REST API Server (Local)

# Set environment variables
export $(cat .env.test | xargs)

# Start the API server
python -m lightrag.api.lightrag_server --host 0.0.0.0 --port 9621

Step 5: Start Web UI (Local)

cd lightrag_webui

# Install dependencies
npm install

# Create .env.local
cat > .env.local << EOF
VITE_API_URL=http://localhost:9621
VITE_ENABLE_AUTH=true
EOF

# Start development server
npm run dev

Test Protocol

Phase 1: Tenant Isolation Tests

Test 1.1: Create Two Tenants via API

# Create Tenant A
curl -X POST http://localhost:9621/api/v1/tenants \
  -H "Content-Type: application/json" \
  -H "X-API-Key: test-api-key-123" \
  -d '{"name": "Tenant Alpha", "description": "Test tenant A"}'

# Create Tenant B
curl -X POST http://localhost:9621/api/v1/tenants \
  -H "Content-Type: application/json" \
  -H "X-API-Key: test-api-key-123" \
  -d '{"name": "Tenant Beta", "description": "Test tenant B"}'

Expected Result: Two distinct tenant IDs returned

Test 1.2: Create KBs per Tenant

# Create KB for Tenant A
curl -X POST http://localhost:9621/api/v1/knowledge-bases \
  -H "Content-Type: application/json" \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_a_id>" \
  -d '{"name": "KB Alpha-1", "description": "Knowledge base for Tenant A"}'

# Create KB for Tenant B
curl -X POST http://localhost:9621/api/v1/knowledge-bases \
  -H "Content-Type: application/json" \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_b_id>" \
  -d '{"name": "KB Beta-1", "description": "Knowledge base for Tenant B"}'

Expected Result: Each KB associated with correct tenant

Test 1.3: Verify KB Isolation

# List KBs for Tenant A (should only see KB Alpha-1)
curl http://localhost:9621/api/v1/knowledge-bases \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_a_id>"

# List KBs for Tenant B (should only see KB Beta-1)
curl http://localhost:9621/api/v1/knowledge-bases \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_b_id>"

Expected Result: Each tenant only sees their own KBs

Phase 2: Document Isolation Tests

Test 2.1: Upload Document to Tenant A

# Upload document
curl -X POST http://localhost:9621/documents/upload \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_a_id>" \
  -H "X-KB-ID: <kb_a_id>" \
  -F "file=@test_doc_a.txt"

Test 2.2: Upload Document to Tenant B

curl -X POST http://localhost:9621/documents/upload \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_b_id>" \
  -H "X-KB-ID: <kb_b_id>" \
  -F "file=@test_doc_b.txt"

Test 2.3: Verify Document Isolation

# List docs for Tenant A (should only see doc A)
curl http://localhost:9621/documents \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_a_id>" \
  -H "X-KB-ID: <kb_a_id>"

# List docs for Tenant B (should only see doc B)
curl http://localhost:9621/documents \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_b_id>" \
  -H "X-KB-ID: <kb_b_id>"

Phase 3: Query Isolation Tests

Test 3.1: Query in Tenant A Context

curl -X POST http://localhost:9621/query \
  -H "Content-Type: application/json" \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_a_id>" \
  -H "X-KB-ID: <kb_a_id>" \
  -d '{"query": "What is in the document?", "mode": "hybrid"}'

Expected Result: Response based only on Tenant A's documents

Test 3.2: Query in Tenant B Context

curl -X POST http://localhost:9621/query \
  -H "Content-Type: application/json" \
  -H "X-API-Key: test-api-key-123" \
  -H "X-Tenant-ID: <tenant_b_id>" \
  -H "X-KB-ID: <kb_b_id>" \
  -d '{"query": "What is in the document?", "mode": "hybrid"}'

Expected Result: Response based only on Tenant B's documents

Phase 4: Web UI Tests

Test 4.1: Tenant Selection

  1. Open Web UI at http://localhost:5173
  2. Login with admin/admin123
  3. Verify tenant selector shows available tenants
  4. Select Tenant A
  5. Verify X-Tenant-ID header is sent with requests

Test 4.2: KB Selection

  1. After selecting Tenant A, verify KB list shows only Tenant A's KBs
  2. Select a KB
  3. Verify X-KB-ID header is sent with requests

Test 4.3: Document Visibility

  1. Select Tenant A / KB Alpha-1
  2. Go to Documents tab
  3. Verify only documents for this KB are shown
  4. Switch to Tenant B / KB Beta-1
  5. Verify documents list changes to show only Tenant B's documents

Test 4.4: Query Scope

  1. Select Tenant A / KB Alpha-1
  2. Enter a query in the chat
  3. Verify response is based on Tenant A's data
  4. Switch to Tenant B
  5. Enter same query
  6. Verify response is different (based on Tenant B's data)

Phase 5: Storage Layer Verification

Test 5.1: Verify PostgreSQL Data Isolation

-- Connect to database
psql -h localhost -U lightrag -d lightrag_test

-- Check tenant_id and kb_id in document tables
SELECT tenant_id, kb_id, id, content_summary 
FROM lightrag_doc_status 
ORDER BY tenant_id, kb_id;

-- Verify no cross-tenant data
SELECT COUNT(*) FROM lightrag_doc_status 
WHERE tenant_id = 'tenant_a' AND kb_id IN (
  SELECT kb_id FROM lightrag_doc_status WHERE tenant_id = 'tenant_b'
);
-- Expected: 0

Test 5.2: Verify Redis Namespace Isolation

redis-cli

# List keys for tenant A
KEYS "tenant_a:*"

# List keys for tenant B  
KEYS "tenant_b:*"

# Verify no overlap

Test Execution Checklist

Phase Test Status Notes
1.1 Create tenants
1.2 Create KBs
1.3 Verify KB isolation
2.1 Upload doc to A
2.2 Upload doc to B
2.3 Verify doc isolation
3.1 Query in A context
3.2 Query in B context
4.1 Tenant selection UI
4.2 KB selection UI
4.3 Document visibility
4.4 Query scope
5.1 PostgreSQL isolation
5.2 Redis isolation

Failure Criteria

A test fails if:

  1. Data from one tenant is visible to another tenant
  2. Query results include information from wrong tenant/KB
  3. Headers are not propagated correctly
  4. Database records show incorrect tenant_id/kb_id associations
  5. Redis keys do not include proper namespace prefixes