Add load testing suite

load-testing/.gitignore

@@ -0,0 +1,17 @@
+# Dependencies
+node_modules/
+
+# Test results
+*.json
+*.csv
+*.log
+
+# OS files
+.DS_Store
+Thumbs.db
+
+# IDE
+.vscode/
+.idea/
+*.swp
+*.swo

load-testing/Makefile

@@ -0,0 +1,23 @@
+.PHONY: help test test-with-monitoring up down clean shell
+
+help: ## Show this help message
+	@echo "Available commands:"
+	@grep -E '^[a-zA-Z_-]+:.*?## .*$$' $(MAKEFILE_LIST) | sort | awk 'BEGIN {FS = ":.*?## "}; {printf "  \033[36m%-15s\033[0m %s\n", $$1, $$2}'
+
+test: ## Run stress-spike test (up to 8000 users, 8 min)
+	docker compose run --rm k6 run tests/main.js
+
+test-with-monitoring: ## Run test with InfluxDB monitoring
+	docker compose run --rm -e K6_OUT=influxdb=http://influxdb:8086 k6 run tests/main.js
+
+up: ## Start InfluxDB and Grafana services
+	docker compose up -d influxdb grafana
+
+down: ## Stop all services
+	docker compose down
+
+clean: ## Stop services and remove volumes
+	docker compose down -v
+
+shell: ## Open shell in k6 container
+	docker compose run --rm k6 sh

load-testing/README.md

@@ -0,0 +1,259 @@
+# Load Testing Suite for dataforgood.fr
+
+A high-intensity load testing suite for the Data for Good website using [Grafana k6](https://k6.io/).
+
+## Overview
+
+This suite tests the performance and scalability of the dataforgood.fr website under intense, realistic traffic conditions. The test simulates thousands of users spending approximately 1 minute on the site, with scenarios including stress testing and traffic spikes.
+
+## Prerequisites
+
+- [Docker](https://docs.docker.com/get-docker/) and Docker Compose
+- (Optional) [Make](https://www.gnu.org/software/make/) for convenient command shortcuts
+
+## Quick Start
+
+### Using Make (Recommended)
+
+```bash
+# Show all available commands
+make help
+
+# Run the stress-spike test
+make test
+
+# Run test with monitoring (stores results in InfluxDB)
+make test-with-monitoring
+
+# Start monitoring stack (InfluxDB + Grafana)
+make up
+
+# Stop all services
+make down
+```
+
+### Using Docker Compose
+
+```bash
+# Run the test
+docker compose run --rm k6 run tests/main.js
+
+# Run with monitoring enabled
+docker compose run --rm -e K6_OUT=influxdb=http://influxdb:8086 k6 run tests/main.js
+
+# Start monitoring stack
+docker compose up -d influxdb grafana
+
+# Access Grafana at http://localhost:3000
+```
+
+## Project Structure
+
+```
+load-testing/
+├── tests/
+│   └── main.js              # Stress-spike test (up to 8000 users)
+├── scenarios/
+│   └── donation-journey.js  # Main page + donation page navigation (~1 min)
+├── utils/
+│   ├── config.js            # Configuration and thresholds
+│   └── helpers.js           # Helper functions
+├── grafana/
+│   └── provisioning/        # Auto-configured datasources and dashboards
+├── docker-compose.yml       # Docker services configuration
+├── Makefile                 # Convenient command shortcuts
+├── package.json             # npm scripts
+└── README.md                # This file
+```
+
+## The Test
+
+### Stress-Spike Test (tests/main.js)
+
+A comprehensive test combining stress testing and traffic spikes to simulate real-world high-traffic scenarios.
+
+**Profile:**
+- **Duration:** 8 minutes
+- **Peak Users:** 8000 concurrent users
+- **User Behaviors:**
+  - 70% complete donation journey (main page → donation page, ~1 min)
+  - 30% quick browsing (main page only, ~30-45 sec)
+
+**Traffic Pattern:**
+1. **Baseline (1 min):** Ramp to 1000 users
+2. **Stress (1 min):** Increase to 3000 users
+3. **Spike (30 sec):** Sudden jump to 8000 users
+4. **Sustained (2 min):** Hold at 8000 users
+5. **Recovery (1 min):** Drop to 3000 users
+6. **Baseline (30 sec):** Return to 1000 users
+7. **Ramp down (1 min):** Complete shutdown
+
+**Run the test:**
+```bash
+make test
+# or with monitoring
+make test-with-monitoring
+```
+
+## User Scenarios
+
+### Donation Journey (~1 minute)
+1. Load the main page (https://dataforgood.fr/)
+2. Scroll/read content (20-30 seconds)
+3. Navigate to donation page (https://dataforgood.fr/faire-un-don)
+4. Read donation page (20-30 seconds)
+
+### Quick Page Load (~30-45 seconds)
+- Load the main page only
+- Quick browsing (30-45 seconds)
+
+## Performance Thresholds
+
+The test uses spike test thresholds to account for extreme load:
+
+- 95th percentile response time < 6000ms (6 seconds)
+- 99th percentile response time < 10000ms (10 seconds)
+- Error rate < 15%
+
+These thresholds are intentionally relaxed to handle the extreme 8000-user spike while still catching major performance degradation.
+
+## Monitoring with Grafana
+
+The suite includes InfluxDB and Grafana services for real-time monitoring.
+
+### Start Monitoring Stack
+
+```bash
+make up
+# or
+docker compose up -d influxdb grafana
+```
+
+### Access Grafana
+
+1. Open http://localhost:3000 in your browser
+2. Login with default credentials (anonymous auth enabled)
+3. The InfluxDB datasource is pre-configured
+
+### Run Tests with InfluxDB Output
+
+```bash
+make test-with-monitoring
+# or
+docker compose run --rm -e K6_OUT=influxdb=http://influxdb:8086 k6 run tests/main.js
+```
+
+## Advanced Usage
+
+### Custom Test Parameters
+
+```bash
+# Run with different target URL
+docker compose run --rm k6 run tests/main.js -e BASE_URL=https://staging.dataforgood.fr
+```
+
+### Output Formats
+
+```bash
+# JSON output
+docker compose run --rm k6 run tests/main.js --out json=results.json
+
+# CSV output
+docker compose run --rm k6 run tests/main.js --out csv=results.csv
+
+# Multiple outputs
+docker compose run --rm k6 run tests/main.js --out json=results.json --out influxdb=http://influxdb:8086
+```
+
+### Interactive Shell
+
+```bash
+make shell
+# or
+docker compose run --rm k6 sh
+```
+
+## Interpreting Results
+
+k6 outputs comprehensive metrics:
+
+- **http_req_duration:** Request latency (p95, p99, median)
+- **http_req_failed:** Percentage of failed requests
+- **http_reqs:** Total HTTP requests per second
+- **vus:** Number of active virtual users
+- **iterations:** Number of scenario completions
+
+Example output:
+```
+     ✓ main page loaded successfully
+     ✓ donation page loaded successfully
+
+     checks.........................: 100.00% ✓ 2340      ✗ 0
+     data_received..................: 45 MB   750 kB/s
+     data_sent......................: 120 kB  2 kB/s
+     http_req_duration..............: avg=234ms min=120ms med=198ms max=890ms p(95)=450ms p(99)=650ms
+     http_reqs......................: 2340    39/s
+     iteration_duration.............: avg=8.2s  min=7.1s  med=8.1s  max=15.8s p(95)=10.2s p(99)=12.5s
+     iterations.....................: 1170    19.5/s
+     vus............................: 50      min=0       max=50
+     vus_max........................: 50      min=50      max=50
+```
+
+## Tips for Success
+
+1. **Start with monitoring:** Always run `make up` first to start Grafana/InfluxDB
+2. **Watch resources:** Monitor server CPU, memory, network, and database during the 8000-user spike
+3. **Establish baselines:** Run tests regularly to track performance trends
+4. **CDN is critical:** With 8000 users, ensure static assets are properly cached
+5. **Database scaling:** Connection pools must handle extreme concurrent load
+6. **Gradual scaling:** Consider starting with lower user counts and adjusting the test stages
+
+## Troubleshooting
+
+### Docker Issues
+
+```bash
+# Restart services
+docker compose restart
+
+# Clean up and rebuild
+docker compose down -v
+docker compose pull
+```
+
+### High Error Rates
+
+- Check server logs for errors
+- Verify the application is running
+- Monitor network connectivity
+- Check if CDN is working properly
+
+### Performance Issues
+
+- Monitor server resources (CPU, memory, disk I/O)
+- Review database query performance
+- Check CDN cache hit rates
+- Analyze slow queries in application logs
+- Monitor network latency
+
+## Cleanup
+
+```bash
+# Stop all services
+make down
+
+# Stop services and remove volumes
+make clean
+
+# Or with docker compose
+docker compose down -v
+```
+
+## References
+
+- [k6 Documentation](https://k6.io/docs/)
+- [k6 with Docker](https://k6.io/docs/getting-started/running-k6/#docker)
+- [Test Types Guide](https://k6.io/docs/test-types/introduction/)
+- [k6 Metrics](https://k6.io/docs/using-k6/metrics/)
+- [Thresholds](https://k6.io/docs/using-k6/thresholds/)
+- [k6 + InfluxDB + Grafana](https://k6.io/docs/results-output/real-time/influxdb/)

load-testing/docker-compose.yml

@@ -0,0 +1,40 @@
+services:
+  k6:
+    image: grafana/k6:latest
+    volumes:
+      - ./:/app
+    working_dir: /app
+    environment:
+      - K6_OUT=${K6_OUT:-influxdb=http://influxdb:8086}
+    command: run tests/main.js
+    depends_on:
+      - influxdb
+
+  # Optional: InfluxDB for storing results
+  influxdb:
+    image: influxdb:1.8
+    ports:
+      - "8086:8086"
+    environment:
+      - INFLUXDB_DB=k6
+      - INFLUXDB_HTTP_AUTH_ENABLED=false
+    volumes:
+      - influxdb-data:/var/lib/influxdb
+
+  # Optional: Grafana for visualizing results
+  grafana:
+    image: grafana/grafana:latest
+    ports:
+      - "3000:3000"
+    environment:
+      - GF_AUTH_ANONYMOUS_ENABLED=true
+      - GF_AUTH_ANONYMOUS_ORG_ROLE=Admin
+    volumes:
+      - grafana-data:/var/lib/grafana
+      - ./grafana/provisioning:/etc/grafana/provisioning
+    depends_on:
+      - influxdb
+
+volumes:
+  influxdb-data:
+  grafana-data:

load-testing/grafana/provisioning/dashboards/dashboard.yml

@@ -0,0 +1,12 @@
+apiVersion: 1
+
+providers:
+  - name: 'k6 Load Testing'
+    orgId: 1
+    folder: ''
+    type: file
+    disableDeletion: false
+    updateIntervalSeconds: 10
+    allowUiUpdates: true
+    options:
+      path: /etc/grafana/provisioning/dashboards

load-testing/grafana/provisioning/datasources/influxdb.yml

@@ -0,0 +1,10 @@
+apiVersion: 1
+
+datasources:
+  - name: InfluxDB_k6
+    type: influxdb
+    access: proxy
+    url: http://influxdb:8086
+    database: k6
+    isDefault: true
+    editable: false