How does caching work and what should I cache first?
I know 'caching makes things fast' but where does one actually start on a real app — browser, CDN, Redis? What's the order?
77
1 answers2.8K views1 Answer
Accepted answer
Caching = keep a copy of expensive-to-compute data somewhere cheap to read. The order that pays off:
1. HTTP/CDN caching (free wins) — static assets with long cache headers; whole pages if they're the same for everyone.
2. Database query results — that homepage query joining five tables and running on every visit? Cache it for 60 seconds in Redis. A 60s TTL on hot reads can cut database load by 90%+.
3. Computed objects — rendered fragments, aggregated counts, expensive API responses.
Two rules that save you pain: every cache entry gets a TTL (no 'cache forever'), and know your invalidation story before you cache — stale-and-eventually-refreshed is fine for a feed, unacceptable for a bank balance.
69
Related questions
- Best free resources to learn system design in 2026?1 answer
- Monolith vs microservices for a small team — what's the honest answer?1 answer
- Horizontal vs vertical scaling — what's the difference and which comes first?2 answers
- What is a load balancer and when do you actually need one?2 answers
- What is the difference between SQL and NoSQL databases, and when do I choose each?1 answer
- How do I start learning system design as a fresher?1 answer