Distributed Cache Design: Redis's Role in System Architecture

Distributed Cache Design

Cache Layer Positions

• CDN: Static assets (images, JS/CSS)
• Application-level cache: In-process LRU cache; fastest but not shared
• Distributed cache: Redis / Memcached; shared across services
• Database query cache: Result-set cache at the DB layer

Redis Data Structures and Use Cases

Structure	Use Cases
String	Session storage, counters, distributed locks
Hash	User profiles, product details
List	Message queues (LPUSH/BRPOP), activity feeds
Set	Tags, deduplication, mutual friends
Sorted Set	Leaderboards, delayed queues

Three Cache Problems

Cache Penetration Querying non-existent keys always hits the DB Fix: Bloom filter to pre-filter / cache null values

Cache Breakdown A hot key expires; massive requests hit DB simultaneously Fix: Mutex lock / logical expiration (no TTL; async refresh)

Cache Avalanche Many keys expire at the same time Fix: Add random TTL jitter / multi-level caching