Performance Testing Quick Reference¶

Quick Start¶

# Run all tests (automated script)
./scripts/run_performance_tests.sh

# Quick test (faster)
./scripts/run_performance_tests.sh --quick

# With load testing
./scripts/run_performance_tests.sh --load

# Everything
./scripts/run_performance_tests.sh --all

Individual Tests¶

Backend Performance Tests¶

cd backend

# All performance tests
poetry run pytest tests/performance/ --performance -v

# Specific test
poetry run pytest tests/performance/test_search_performance.py::TestSearchPerformance::test_semantic_search_latency --performance -v

Load Testing¶

cd backend/tests/performance

# Install dependencies (first time only)
pip install -r requirements.txt

# Interactive mode (web UI at http://localhost:8089)
locust -f locustfile.py --host=http://localhost:8004

# Headless mode
locust -f locustfile.py --host=http://localhost:8004 --users 50 --spawn-rate 5 --run-time 5m --headless --html=report.html

Performance Monitor¶

# Quick benchmark
python scripts/performance_monitor.py

# Save results
python scripts/performance_monitor.py --output results.json

Frontend Analysis¶

cd frontend
node scripts/analyze-bundle.js

Performance Targets¶

Metric	Target	Acceptable
Health Check p95	50ms	100ms
Search p95	2s	3s
Vector Search p95	500ms	1s
Search Throughput	10 req/s	5 req/s
Success Rate	>99%	>95%

Common Commands¶

# Backend server (required for tests)
cd backend
poetry run uvicorn app.server:app --port 8004

# Run quick performance check
./scripts/run_performance_tests.sh --quick --monitor

# Run full suite
./scripts/run_performance_tests.sh --all

# Load test - light
locust -f backend/tests/performance/locustfile.py --host=http://localhost:8004 --users 10 --spawn-rate 2 --run-time 2m --headless

# Load test - heavy
locust -f backend/tests/performance/locustfile.py --host=http://localhost:8004 --users 200 --spawn-rate 10 --run-time 10m --headless

Interpreting Results¶

Latency Metrics¶

p50: Median (50% faster)
p95: 95th percentile ⭐ KEY METRIC
p99: 99th percentile (worst case)

Success Metrics¶

Success Rate: % of requests that succeed
Throughput: Requests per second
Error Rate: % of failed requests

Troubleshooting¶

# Tests skipped?
# → Add --performance flag
poetry run pytest tests/performance/ --performance

# Backend not running?
# → Start backend server
cd backend && poetry run uvicorn app.server:app --port 8004

# Locust not installed?
# → Install dependencies
cd backend/tests/performance && pip install -r requirements.txt

# Import errors?
# → Check PYTHONPATH and install packages
cd backend && poetry install

Files¶

Tests: backend/tests/performance/
Scripts: scripts/performance_monitor.py
Docs: docs/PERFORMANCE_TESTING.md
Summary: PERFORMANCE_SUMMARY.md