Redis Best Practices

20. Redis Best Practices

7 Pillars of Redis Best Practices

flowchart TB
    Pillars[Redis Best Practices 🏆]

    Pillars --> P1[1. Key Naming<br/>user:42:name<br/>Colons as separators]
    Pillars --> P2[2. Always Set TTL<br/>EX / EXPIRE<br/>Prevent memory leaks]
    Pillars --> P3[3. Memory Optimization<br/>Use Hashes over<br/>multiple keys]
    Pillars --> P4[4. Pipelining<br/>Batch commands<br/>Reduce round trips]
    Pillars --> P5[5. Connection Pooling<br/>ioredis built-in<br/>Reuse connections]
    Pillars --> P6[6. Monitor Production<br/>INFO / SLOWLOG / MEMORY
Watch your metrics]
    Pillars --> P7[7. Separate Instances<br/>Session / Cache / Queue<br/>Isolate by purpose]

    style Pillars fill:#7c3aed,color:#fff
    style P1 fill:#3b82f6,color:#fff
    style P2 fill:#059669,color:#fff
    style P3 fill:#f59e0b,color:#fff
    style P4 fill:#ec4899,color:#fff
    style P5 fill:#06b6d4,color:#fff
    style P6 fill:#10b981,color:#fff
    style P7 fill:#8b5cf6,color:#fff

Golden rule: Redis is an in-memory cache, not a primary database. Design your architecture around that.

1. Key Naming Conventions

Use a consistent hierarchy separated by colons (:):

{app}:{resource}:{id}:{attribute}

Examples:
  myapp:user:42:session
  myapp:product:SKU-9001:cache
  myapp:ratelimit:user:42:2024-01-15
  myapp:leaderboard:game:season3

Rules:

Use lowercase
Use colons as separators
Be descriptive but concise
Include version if needed: cache:products:v2

2. Always Set TTL on Cache Keys

// Set TTL based on data volatility:
// Frequently changing data (prices, stock) → 60–300 seconds
// Stable data (product descriptions) → 3600–86400 seconds
// Session data → match your session timeout

redis.setex('cache:prices', 60, JSON.stringify(prices));
redis.setex('cache:products', 3600, JSON.stringify(products));
redis.setex('session:user:42', 86400, token);

3. Memory Optimization

# Use Hash for objects with many fields (more memory-efficient than multiple keys)
# ❌ Inefficient — 4 separate keys
SET user:42:name "Alice"
SET user:42:age "30"
SET user:42:city "Pune"
SET user:42:role "admin"

# ✅ Efficient — 1 Hash with 4 fields
HSET user:42 name "Alice" age "30" city "Pune" role "admin"

4. Use Pipelining for Bulk Operations

// ❌ BAD — 1000 round trips to Redis
for (const user of users) {
  await redis.set(`user:${user.id}`, JSON.stringify(user));
}

// ✅ GOOD — 1 round trip with pipeline
const pipeline = redis.pipeline();
for (const user of users) {
  pipeline.set(`user:${user.id}`, JSON.stringify(user), 'EX', 3600);
}
await pipeline.exec();

5. Connection Pooling

// ioredis handles connection pooling internally
// For cluster mode:
const Redis = require('ioredis');
const cluster = new Redis.Cluster([
  { host: '127.0.0.1', port: 7000 },
  { host: '127.0.0.1', port: 7001 },
  { host: '127.0.0.1', port: 7002 },
]);

6. Monitor Redis in Production

# Real-time command monitoring (development only)
redis-cli MONITOR

# Check memory usage
redis-cli INFO memory

# Check connected clients
redis-cli INFO clients

# Check slow queries (commands taking > 10ms)
redis-cli SLOWLOG GET 10

# Memory usage of a specific key
redis-cli MEMORY USAGE user:42

7. Separate Redis Instances by Purpose

Redis Instance 1 → Session storage (db: 0)
Redis Instance 2 → API cache (db: 0)
Redis Instance 3 → Rate limiting (db: 0)

Or use Redis databases (0–15) for logical separation:

const sessionRedis = new Redis({ db: 0 });
const cacheRedis   = new Redis({ db: 1 });
const queueRedis   = new Redis({ db: 2 });