Chapter 01 · The slow path
Every read hits the database
Alice, Bob and Carol keep asking for the same few keys.
With no cache, every ask goes to the database, including a key that was just read.
System design · interactive
Watch a small cache hit, miss and forget a key. Three moments stop and ask you to predict before the picture answers.
1 / 10
Hits
0
Hit rate
0%
Database
slow
“a” is read from the database. The same keys will be computed again.
Chapter 01 · The slow path
Alice, Bob and Carol keep asking for the same few keys.
With no cache, every ask goes to the database, including a key that was just read.
In production
The same hit / miss / eviction ideas show up in dedicated cache servers and in front of your app — in any language.
Not tied to one language — use these names when you search docs or design a system.
In-memory key–value store with TTL, LRU-style eviction policies, and replication.
Official site ↗Simple, fast distributed memory cache for hot keys.
Official site ↗Managed Redis or Memcached on AWS.
Official site ↗Cloudflare, Fastly, Akamai — cache HTTP responses close to users.
Official site ↗Browser and proxy caching without a custom app cache.
Documentation ↗