Caching Strategies Cheat Sheet

Cache patterns and eviction policies — cache-aside, read-through, write-through, write-behind, Redis, Memcached, and CDN caching.

Last Updated: May 1, 2025

Cache Patterns

Pattern	Read Path	Write Path
Cache-Aside	Check cache → miss → DB → populate	Invalidate after DB write
Read-Through	Cache handles DB reads	N/A (writes direct to DB)
Write-Through	Cache always fresh	Write cache → cache writes DB
Write-Behind	Same as cache-aside	Write cache → async DB write

Eviction Policies

Policy	How It Works	Best For
LRU	Evict least recently used	General purpose — great default
LFU	Evict least frequently used	Hot items that stay hot
TTL	Evict after fixed time	Sessions, rate limits, OTPs
FIFO	Evict oldest by insertion	Simple, bounded memory

Redis Caching Commands

SET key value EX 3600

Set key with 1-hour TTL

GET key

Retrieve value by key

DEL key

Delete a key (manual invalidation)

EXPIRE key 300

Set/update TTL to 5 minutes

MGET key1 key2

Get multiple keys in one round trip

INCR key

Atomic increment — counters, rate limits

Cache Invalidation

Item	Description
`TTL-Based`	Simplest — data expires after fixed time
`Write Invalidation`	Delete cache entry after DB write
`Event-Driven`	Publish change events → cache subscribers invalidate
`Versioned Keys`	Append version: user:123:v2 — no invalidation needed

Pro Tip: There are only two hard things in Computer Science: cache invalidation and naming things. Always define a TTL and invalidation strategy before adding a cache.