| name | nestjs-caching |
| description | Implement multi-level caching, invalidation patterns, and stampede protection in NestJS. Use when adding Redis caching layers, configuring cache-manager interceptors, implementing stale-while-revalidate, or preventing cache stampedes in NestJS services. |
| metadata | {"triggers":{"files":["**/*.service.ts","**/*.interceptor.ts"],"keywords":["CacheInterceptor","CacheTTL","Redis","stale-while-revalidate"]}} |
Caching & Redis Standards
Priority: P1 (HIGH)
Caching Strategy
- Layering: Use Multi-Level Caching for high-traffic read endpoints.
- L1 (Local): In-Memory (Node.js heap). Ultra-fast, no network. Use
lru-cache for config/static data.
- L2 (Distributed): Redis. Shared across pods.
- Pattern: Implement Stale-While-Revalidate to avoid latency spikes during cache misses.
NestJS Implementation
- Library: Use
cache-manager with cache-manager-redis-yet (recommended over cache-manager-redis-store for V4 stability).
- Interceptors: Use
@UseInterceptors(CacheInterceptor) for simple GET responses.
- Warning: Default key URL. Ensure consistent query param ordering or use custom key generators.
See implementation examples
Stampede Protection
- Jitter: Add random variance to TTLs to prevent simultaneous expiry across keys.
- Locking: One process recomputes while others wait or return stale data.
See implementation examples
Redis Data Structures
- Hash (
HSET): Store objects (user profiles) with partial update support.
- Set (
SADD): Unique collections with O(1) membership checks.
- Sorted Set (
ZADD): Priority queues, leaderboards, rate-limiting windows.
Invalidation Patterns
- Tagging: Use Sets to group cache keys (avoid
KEYS which O(N) in production).
- Create:
SADD post:1:tags cache:post:1
- Invalidate: Fetch tags from Set, then
DEL grouped keys.
- Event-Driven: Listen to domain events (
UserUpdated) to trigger invalidation asynchronously.
Anti-Patterns
- No KEYS in production: Use SET-based tag grouping for cache invalidation; KEYS O(N).
- No fixed TTLs on grouped caches: Add jitter (±10s) to prevent simultaneous stampede.
- No MemoryStorage for multi-pod: Use Redis store; in-memory cache not shared across pods.
