Reqflow
← All learning paths
Learning path
Beginner~55 min·3 systems

Caching & Read-Heavy Systems

How read-dominated systems use caches, CDNs, and precomputation to make the hot path effectively free.

After this path you will be able to

Design any read-heavy system: identify where to add a cache, choose between CDN and in-process caching, and explain the cache invalidation trade-offs to an interviewer.

Interview approach for this path
  1. 1.Start by asking: is this a read-heavy or write-heavy system? Caching only helps reads.
  2. 2.Clarify the cache hit rate you'd need. Below 80% and caching may not be worth the added complexity.
  3. 3.Name the specific caching strategy: cache-aside, write-through, or write-back, and explain why you chose it.
  4. 4.Call out the eviction policy: LRU for general use, LFU for skewed access patterns.
  5. 5.Talk about cache invalidation explicitly. This is where most follow-up questions come from.
  6. 6.Mention thundering herd and how you'd protect a cold cache on startup or after a node failure.
  7. 7.Address CDN placement for static and semi-static assets before the interviewer brings it up.

Systems in this path

3 total
  1. 1
    URL Shortener
    Beginner·12 min

    Hashing, key generation, read-heavy caching.

  2. 2
    Instagram Feed
    Intermediate·18 min

    Fan-out on write vs read, ranking, CDN.

  3. 3
    Netflix (Video Streaming)
    Advanced·25 min

    CDN-first, adaptive bitrate, recommendation.

Concepts reinforced throughout

Up next

Async Patterns

Queues, fan-out, retries, webhooks: when and why to decouple work from the request path.