Aller au contenu principal

Caching System

Learn how to leverage the powerful Redis caching layer for ultra-fast API responses.

Overview​

Express Base API includes a sophisticated caching system that:

  • ⚡ Reduces response time by 50-100x
  • 🔄 Dual backend - Redis (primary) with in-memory fallback
  • 🎯 Automatic invalidation on data changes
  • 📊 Built-in statistics to monitor performance
  • 🛡️ Type-safe with full TypeScript support

Architecture​

Request → Cache Check → Cache Hit? → Response (< 1ms)
                     ↓
                  Cache Miss
                     ↓
               Database Query → Cache Result → Response

Basic Usage​

In Services​

All services extending BaseService have caching methods available:

import BaseService from '../../../utils/bases/base.service';

class ProductService extends BaseService<IProduct> {
  async getProduct(id: string) {
    // Find by ID with automatic caching (1 hour TTL)
    return await this.findByIdCached(id, {}, null, 3600);
  }

  async getFeaturedProducts() {
    // Cache-aside pattern
    return await this.cacheGetOrSet(
      'products:featured',
      async () => {
        return await this.find(
          { featured: true } as any,
          { sort: { views: -1 }, limit: 10 }
        );
      },
      3600 // 1 hour TTL
    );
  }
}

Manual Caching​

// Set cache
await service.cacheSet('user:123', userData, 3600); // 1 hour

// Get from cache
const user = await service.cacheGet('user:123');

// Delete from cache
await service.cacheDelete('user:123');

// Delete by pattern
await service.cacheDeletePattern('user:*');

Cache Methods Reference​

Protected Methods (available in services)​

cacheGet<T>(key: string): Promise<T | null>​

Get value from cache.

const product = await this.cacheGet<IProduct>('product:123');
if (product) {
  return { error: false, data: product };
}

cacheSet(key: string, value: any, ttl: number): Promise<boolean>​

Store value in cache with TTL (Time-To-Live) in seconds.

await this.cacheSet('product:123', productData, 3600); // 1 hour

cacheDelete(key: string): Promise<boolean>​

Remove specific cache entry.

await this.cacheDelete('product:123');

cacheDeletePattern(pattern: string): Promise<number>​

Remove multiple entries matching a pattern.

const deleted = await this.cacheDeletePattern('product:*');
console.log(`Deleted ${deleted} cache entries`);

cacheGetOrSet<T>(key: string, fetcher: Function, ttl: number): Promise<T | null>​

Cache-aside pattern - get from cache or execute fetcher and cache result.

const products = await this.cacheGetOrSet(
  'products:category:electronics',
  async () => await this.find({ category: 'electronics' } as any),
  3600
);

Public Methods​

findByIdCached(id, query, model, ttl): Promise<QueryResult<T>>​

Find by ID with automatic caching.

const result = await productService.findByIdCached(
  '123',
  { populate: 'category' },
  null,
  3600
);

invalidateCache(id, model): Promise<boolean>​

Invalidate cache for a specific document.

await productService.update(productId, updateData);
await productService.invalidateCache(productId);

invalidateAllCache(model): Promise<number>​

Invalidate all cache entries for a model.

const deleted = await productService.invalidateAllCache();

Authentication Caching​

The authentication system automatically caches user data for blazing-fast subsequent requests.

How It Works​

  1. On Login - User + Auth data cached (1-hour TTL)
  2. On Authentication - Middleware checks cache first (< 1ms)
  3. On Logout/Password Reset - Cache automatically invalidated

Performance Impact​

OperationWithout CacheWith CacheImprovement
Auth check50-100ms< 1ms50-100x faster
API requestDB query each timeCache hitMassive DB load reduction

Example Flow​

// 1. User logs in
POST /auth/login
// → User data cached for 1 hour (key: user:auth:{userId})

// 2. Subsequent authenticated requests
GET /api/profile
// → Middleware checks cache (< 1ms)
// → No DB query needed! ✅

// 3. User logs out
POST /auth/logout
// → Cache invalidated
// → Next request requires DB query

Cache Statistics​

Monitor cache performance:

import cacheService from './utils/helpers/cache.helper';

const stats = cacheService.getStats();
console.log(stats);

Output:

{
  "hits": 1500,
  "misses": 250,
  "sets": 250,
  "deletes": 50,
  "errors": 0,
  "hitRate": "85.71%",
  "backend": "redis",
  "size": undefined
}

Reset Statistics​

cacheService.resetStats();

Best Practices​

1. Choose Appropriate TTL​

// Short-lived data (user session)
await this.cacheSet('session:123', session, 3600); // 1 hour

// Medium-lived data (product listings)
await this.cacheSet('products:featured', products, 86400); // 1 day

// Long-lived data (static content)
await this.cacheSet('config:site', config, 604800); // 1 week

2. Use Descriptive Keys​

// ✅ Good - Clear and hierarchical
'user:123'
'user:123:profile'
'products:category:electronics'
'products:featured:2024-01-15'

// ❌ Bad - Unclear and hard to invalidate
'u123'
'temp_data'
'cache_12345'

3. Invalidate on Updates​

async updateProduct(id: string, data: any) {
  // Update database
  const result = await this.update({ _id: id } as any, data);

  if (!result.error) {
    // Invalidate specific cache
    await this.invalidateCache(id);

    // Invalidate related caches
    await this.cacheDelete('products:featured');
    await this.cacheDeletePattern(`products:category:*`);
  }

  return result;
}

4. Handle Cache Miss Gracefully​

async getProduct(id: string) {
  // Try cache first
  const cached = await this.cacheGet<IProduct>(`product:${id}`);
  if (cached) {
    return { error: false, data: cached };
  }

  // Cache miss - query database
  const result = await this.findById(id);

  // Cache for next time
  if (!result.error && result.data) {
    await this.cacheSet(`product:${id}`, result.data, 3600);
  }

  return result;
}

Advanced Patterns​

Cache Warming​

Pre-populate cache with frequently accessed data:

async warmCache() {
  // Get most popular products
  const popular = await this.find(
    {} as any,
    { sort: { views: -1 }, limit: 100 }
  );

  // Cache each product
  if (!popular.error && popular.data) {
    for (const product of popular.data) {
      await this.cacheSet(
        `product:${product._id}`,
        product,
        3600
      );
    }
  }
}

Conditional Caching​

Cache based on conditions:

async getProducts(filters: any, useCache: boolean = true) {
  const cacheKey = `products:${JSON.stringify(filters)}`;

  if (useCache) {
    const cached = await this.cacheGet(cacheKey);
    if (cached) return { error: false, data: cached };
  }

  const result = await this.find(filters);

  if (useCache && !result.error) {
    await this.cacheSet(cacheKey, result.data, 3600);
  }

  return result;
}

Cache Tags​

Invalidate multiple related caches:

async createProduct(data: any) {
  const result = await this.create(data);

  if (!result.error) {
    // Invalidate all product-related caches
    await this.cacheDeletePattern('products:*');
    await this.cacheDeletePattern(`category:${data.category}:*`);
  }

  return result;
}

Troubleshooting​

Cache Not Working​

  1. Check Redis connection:

    redis-cli ping
    # Should respond: PONG

  2. Verify configuration:

    # In .env
    USE_QUEUE=yes
    REDIS_HOST=localhost
    REDIS_PORT=6379

  3. Check logs:

    ✅ Cache service using Redis
    # or
    ⚠️ Cache service falling back to in-memory cache

Low Hit Rate​

  • TTL too short - increase cache duration
  • Keys changing too often - use stable cache keys
  • Cache invalidated too aggressively - review invalidation logic

Memory Issues​

Monitor Redis memory:

redis-cli INFO memory

Set max memory policy in Redis config:

maxmemory 256mb
maxmemory-policy allkeys-lru

Configuration​

Redis Configuration​

configs/redis.config.ts
const redisOptions = {
  host: configs.getValue('REDIS_HOST'),
  port: Number(configs.getValue('REDIS_PORT')),
  password: configs.getValue('REDIS_PASSWORD', false),
  maxRetriesPerRequest: 3,
  retryStrategy: (times) => Math.min(times * 50, 2000),
};

Cache Service Configuration​

Default TTL and other settings can be customized in utils/helpers/cache.helper.ts.

Next Steps​

Questions? Check the GitHub Discussions or open an issue.