Introduction

Performance is product. Here’s our battle‑tested checklist for Rails apps that want LCP ≤ 2.5s, INP ≤ 200ms, and CLS ≤ 0.1—without guesswork.

1) Measure First

• CrUX & PageSpeed field data for key pages
• APM traces (New Relic/Datadog): p95 latency per endpoint
• Error rate and slow queries; sample traces

2) Database

• Add missing/composite indexes; remove duplicates
• Kill N+1s (bullet, rack-mini-profiler)
• Partition/archive hot tables; tune autovacuum
• Read replicas carefully—watch staleness

3) Caching & Assets

• Page/fragment caching with cache keys (avoid time‑based expiry)
• Redis TTLs & memory policy sanity check
• CDN for static assets; correct Cache-Control & ETag

4) Background Work

• Move heavy work off request cycle (Sidekiq)
• Idempotent jobs with retries/backoff
• Dedicated queues for critical paths

5) Frontend Vitals

• Inline critical CSS; self‑host fonts with font-display: swap
• Image width/height + lazy loading (except hero); use next/image
• Break up long tasks; avoid hydrating huge bundles

6) Release Engineering

• CI < 10 min with test splitting & caching
• Canary deploys + feature flags + error budgets
• Build once; promote between envs

A Minimal NGINX/Cache Header Tweak

location ~* \.(js|css|png|webp|jpg|svg)$ {
  add_header Cache-Control "public, max-age=31536000, immutable";
  try_files $uri =404;
}
location / {
  add_header Cache-Control "no-store";
}

30‑Point Checklist

• All hot queries have indexes; N+1s fixed
• Redis keys scoped; eviction policy appropriate
• Hero images ≤ 120 KB, WebP/AVIF used
• Fonts self‑hosted; display: swap
• Core Web Vitals monitored in RUM (p75)
• Canary deploy path + rollback verified
• APM traces sampled with error budgets
• Queues sized for peak; dead letters empty

Conclusion

Performance isn’t a sprint; it’s a habit. Start with measurement, fix the biggest regressions, and guard the wins in CI/CD and observability.

Written by Ahmad
Ahmad Hassan, Senior Ruby on Rails expert with 10+ years of experience. He builds web and mobile applications using React, Next.js, and TypeScript, and contributes to backend development using Python and Ruby on Rails.