Performance Tuning Guide

Version: 1.0.0 | Target: P99 < 5ms, Throughput > 100k req/s

Current Performance (Oct 2025)

Latency P99: 752ms (1 vCPU, sustained load)
Throughput: 287 req/s
Memory: ~80MB
CO2/request: 0.12g

Backend Optimization

1. Database Connection Pool

// backend/src/main.rs
let pool = PgPoolOptions::new()
    .max_connections(20)        // Increase from 10
    .acquire_timeout(Duration::from_secs(3))
    .connect(&database_url)
    .await?;

2. Query Optimization

-- Add indexes for frequent queries
CREATE INDEX idx_expenses_building_status ON expenses(building_id, status);
CREATE INDEX idx_unit_owners_unit_active ON unit_owners(unit_id) WHERE end_date IS NULL;

-- Analyze query plans
EXPLAIN ANALYZE SELECT * FROM expenses WHERE building_id = $1 AND status = 'Approved';

3. Caching Strategy

use once_cell::sync::Lazy;
use std::collections::HashMap;

static ACCOUNT_CACHE: Lazy<Mutex<HashMap<String, Account>>> = Lazy::new(|| {
    Mutex::new(HashMap::new())
});

4. Async Optimization

// Use join instead of sequential awaits
let (buildings, units, owners) = tokio::join!(
    fetch_buildings(),
    fetch_units(),
    fetch_owners()
);

Database Tuning

PostgreSQL Configuration

# /etc/postgresql/15/main/postgresql.conf

shared_buffers = 256MB              # 25% of RAM
effective_cache_size = 768MB        # 75% of RAM
maintenance_work_mem = 64MB
checkpoint_completion_target = 0.9
wal_buffers = 16MB
default_statistics_target = 100
random_page_cost = 1.1              # SSD
effective_io_concurrency = 200      # SSD
work_mem = 4MB
min_wal_size = 1GB
max_wal_size = 4GB
max_worker_processes = 2
max_parallel_workers_per_gather = 1
max_parallel_workers = 2

Frontend Optimization

1. Code Splitting

---
// Lazy load heavy components
const HeavyChart = () => import('../components/HeavyChart.svelte');
---

2. Image Optimization

# Convert to WebP
npm install sharp
npx sharp input.jpg -o output.webp --webp

# Serve with <picture>
<picture>
  <source srcset="image.webp" type="image/webp">
  <img src="image.jpg" alt="...">
</picture>

3. Bundle Analysis

npm run build
npx vite-bundle-visualizer

Monitoring & Profiling

Backend Profiling

# Install flamegraph
cargo install flamegraph

# Profile CPU
sudo cargo flamegraph --bin koprogo-api

# Profile memory
valgrind --tool=massif ./target/release/koprogo-api

Database Profiling

-- Enable pg_stat_statements
CREATE EXTENSION IF NOT EXISTS pg_stat_statements;

-- Find slow queries
SELECT query, mean_exec_time, calls
FROM pg_stat_statements
ORDER BY mean_exec_time DESC
LIMIT 10;

Load Testing

cd load-tests
k6 run --vus 100 --duration 60s scenarios/api-load.js

Quick Wins

✅ Enable gzip compression (Nginx/Traefik)
✅ Use CDN for static assets
✅ Add database indexes on foreign keys
✅ Cache PCMN accounts (rarely change)
✅ Use connection pooling
✅ Implement HTTP caching headers

Version: 1.0.0