@unidev-hub/cache

A unified caching layer for Node.js applications that provides a consistent interface to different caching backends with advanced features.

Features

• 🚀 Multiple Cache Providers

  • In-memory caching for single-instance applications
  • Redis-based caching for distributed applications
  • Tiered caching for optimal performance (e.g., memory L1 + Redis L2)

• 💪 Advanced Caching Strategies

  • TTL-based expiration
  • LRU/LFU eviction policies
  • Write-through, write-behind, and write-around strategies

• 🔄 Cache Invalidation

  • Pattern-based invalidation (supports wildcards)
  • Tag-based invalidation for related items
  • Version-based invalidation (for bulk invalidation)

• 🔒 Distributed Locks

  • Memory-based locks for single-instance applications
  • Redis-based locks for distributed coordination
  • Auto-extending locks to prevent expiration during long operations

• ⚙️ Flexible Configuration

  • Simple API with sensible defaults
  • Highly customizable caching behavior
  • Built-in logging and monitoring

Installation

npm install @unidev-hub/cache

or

yarn add @unidev-hub/cache

Quick Start

import { CacheManager } from '@unidev-hub/cache';

// Create an in-memory cache
const cache = new CacheManager();

// Basic operations
await cache.set('key', { hello: 'world' });
const value = await cache.get('key');
console.log(value); // { hello: 'world' }

// With TTL (time-to-live)
await cache.set('temporary', 'expires soon', 5000); // 5 seconds

// Cache or compute pattern
const expensiveValue = await cache.getOrSet(
  'expensive-operation',
  async () => {
    // This function only runs if the key isn't in the cache
    console.log('Computing expensive value...');
    return await someExpensiveOperation();
  },
  3600000 // 1 hour TTL
);

Cache Providers

In-Memory Cache

The default provider is an in-memory cache, suitable for single-instance applications or testing.

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager({
  memory: {
    maxItems: 10000,   // Maximum number of items to store
    maxSize: 100000000, // Maximum cache size in bytes (100MB)
    evictionStrategy: 'lru', // 'lru', 'lfu', or 'fifo'
    checkExpirationInterval: 60000, // Check for expired items every 1 minute
  }
});

Redis Cache

For distributed applications, use the Redis cache provider:

import { CacheManager } from '@unidev-hub/cache';
import Redis from 'ioredis';

const redis = new Redis({
  host: 'localhost',
  port: 6379,
  // other Redis options
});

const cache = new CacheManager({
  redis: {
    client: redis,    // Existing Redis client
    // Or connection options:
    // connection: {
    //   host: 'localhost',
    //   port: 6379,
    //   password: 'secret',
    //   db: 0
    // },
    ttl: 3600000,     // Default TTL (1 hour)
    keyPrefix: 'app:cache:',  // Prefix for all keys
  }
});

Tiered Cache

Combine multiple cache providers in a tiered architecture for optimal performance:

import { CacheManager, MemoryCache, RedisCache } from '@unidev-hub/cache';
import Redis from 'ioredis';

const redis = new Redis();

const cache = new CacheManager({
  tiered: {
    providers: [
      {
        provider: new MemoryCache({ maxItems: 10000 }),
        options: {
          writeOnHitFromLowerTier: true,
          ttl: 300000, // 5 minutes for L1 (memory)
        }
      },
      {
        provider: new RedisCache({ client: redis }),
        options: {
          ttl: 3600000, // 1 hour for L2 (Redis)
        }
      }
    ],
    writeStrategy: 'write-through', // 'write-through', 'write-behind', or 'write-around'
    propagateDeletes: true,   // Delete from all tiers
    propagateUpdates: true,   // Update all tiers
  }
});

Cache Invalidation

Pattern-based Invalidation

Invalidate cache entries using wildcard patterns:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager({
  invalidation: {
    method: 'pattern',
    maxKeys: 1000, // Max keys to invalidate in one operation
  }
});

// Set some values
await cache.set('user:1:profile', { name: 'Alice' });
await cache.set('user:1:settings', { theme: 'dark' });
await cache.set('user:2:profile', { name: 'Bob' });

// Invalidate all keys for user 1
const invalidatedCount = await cache.invalidate('user:1:*');
console.log(`Invalidated ${invalidatedCount} entries`); // 2

Tag-based Invalidation

Associate cache entries with tags and invalidate by tag:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager({
  invalidation: {
    method: 'tag',
  }
});

// Set values with tags
await cache.set('user:1', { name: 'Alice' }, 3600000, {
  method: 'tag',
  tags: ['users', 'user-1']
});

await cache.set('user:2', { name: 'Bob' }, 3600000, {
  method: 'tag',
  tags: ['users', 'user-2']
});

// Invalidate by a single tag
await cache.invalidate(undefined, ['users']); // All users

// Invalidate by multiple tags (AND operation)
await cache.invalidate(undefined, ['users', 'vip']); // Users who are VIPs

Version-based Invalidation

Namespace your cache keys with versions for bulk invalidation:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager({
  invalidation: {
    method: 'version',
  }
});

// Set values in a namespace
await cache.set('product:1', { name: 'Laptop' }, 3600000, {
  method: 'version',
  namespace: 'products'
});

await cache.set('product:2', { name: 'Phone' }, 3600000, {
  method: 'version',
  namespace: 'products'
});

// Invalidate an entire namespace by incrementing its version
await cache.invalidate(undefined, undefined, 'products');

Distributed Locks

Basic Lock Usage

Acquire locks to ensure exclusive access to resources:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager({
  redis: {
    // Redis configuration
  }
});

// Get the lock manager
const lockManager = cache.getLockManager();

// Acquire a lock
const lock = await lockManager.acquire(
  'resource-name',
  10000, // TTL: 10 seconds
  {
    wait: true, // Wait if lock is held
    retryDelay: 200, // 200ms between retries
    maxRetries: 50,  // Max retry attempts
    autoExtend: true, // Auto-extend the lock while held
  }
);

if (lock) {
  try {
    // Perform operations with exclusive access
    await someOperation();
  } finally {
    // Always release the lock when done
    await lockManager.release(lock);
  }
} else {
  console.log('Failed to acquire lock');
}

Using withLock Helper

Simplify locking with the withLock helper:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager();
const lockManager = cache.getLockManager();

// Execute function with automatic lock management
const result = await lockManager.withLock(
  'resource-name',
  async (lock) => {
    // This code runs with exclusive access
    return await someOperation();
  },
  10000, // TTL: 10 seconds
  {
    wait: true,
    maxRetries: 50,
  }
);

Caching with Locking

Use wrapWithLock to combine caching and locking:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager();

// Create a function that caches results and ensures exclusive access
const getCachedUserWithLock = cache.wrapWithLock(
  async (userId) => {
    // This expensive operation runs with exclusive access
    // and only when the result isn't cached
    return await fetchUserFromDatabase(userId);
  },
  (userId) => `user:${userId}`, // Key generator
  { ttl: 5000, wait: true }, // Lock options
  3600000, // Cache TTL: 1 hour
  { method: 'tag', tags: ['users'] } // Invalidation options
);

// Usage
const user = await getCachedUserWithLock(123);

Advanced Usage

Function Wrapping

Automatically cache function results:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager();

// Wrap a function with caching
const getCachedUser = cache.wrap(
  async (userId) => {
    console.log('Fetching user from database...');
    return await fetchUserFromDatabase(userId);
  },
  (userId) => `user:${userId}`, // Key generator function
  3600000 // 1 hour TTL
);

// First call executes the function
const user1 = await getCachedUser(123);

// Second call retrieves from cache
const user2 = await getCachedUser(123);

Custom Key Generation

Generate consistent cache keys for complex objects:

import { CacheManager } from '@unidev-hub/cache';

// Generate a cache key from an object
const key = CacheManager.generateKey({
  userId: 123,
  filters: { status: 'active', role: 'admin' },
  sort: { field: 'created', order: 'desc' }
}, {
  prefix: 'users:list:',
  includeType: true,
  hash: true, // Hash the key for length consistency
});

// Use the generated key
const cache = new CacheManager();
await cache.set(key, resultData);

Batch Operations

Perform operations on multiple keys at once:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager();

// Set multiple values
const values = new Map();
values.set('key1', 'value1');
values.set('key2', 'value2');
values.set('key3', 'value3');
await cache.setMany(values, 3600000);

// Get multiple values
const keys = ['key1', 'key2', 'key4'];
const results = await cache.getMany(keys);
console.log(results.get('key1')); // 'value1'
console.log(results.has('key4')); // false (not found)

// Delete multiple values
await cache.deleteMany(['key1', 'key2']);

Monitoring and Statistics

Get cache statistics for monitoring:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager();

// Get cache statistics
const stats = await cache.getStats();
console.log(`Cache size: ${stats.size} items`);
console.log(`Hit ratio: ${stats.hitRatio * 100}%`);
console.log(`Hits: ${stats.hits}, Misses: ${stats.misses}`);

Logging

Enable logging to debug cache operations:

import { CacheManager } from '@unidev-hub/cache';

const cache = new CacheManager({
  enableLogging: true,
  logLevel: 'debug', // 'debug', 'info', 'warn', or 'error'
});

Configuration Reference

Global Options

interface CacheOptions {
  // Default TTL for cache entries in milliseconds
  ttl?: number; // Default: 3600000 (1 hour)
  
  // Key prefix for all cache entries
  keyPrefix?: string; // Default: 'cache:'
  
  // Whether to serialize values (helps with complex objects)
  serialize?: boolean; // Default: true
  
  // Logging options
  enableLogging?: boolean; // Default: false
  logLevel?: 'debug' | 'info' | 'warn' | 'error'; // Default: 'info'
  
  // Provider-specific options
  memory?: MemoryCacheOptions;
  redis?: RedisCacheOptions;
  tiered?: TieredCacheOptions;
  
  // Invalidation options
  invalidation?: InvalidationOptions;
}

Memory Cache Options

interface MemoryCacheOptions {
  // Maximum number of items in the cache
  maxItems?: number; // Default: 1000
  
  // Maximum size of the cache in bytes
  maxSize?: number;
  
  // Eviction strategy when cache is full
  evictionStrategy?: 'lru' | 'lfu' | 'fifo'; // Default: 'lru'
  
  // Check interval for expired items in milliseconds
  checkExpirationInterval?: number; // Default: 60000 (1 minute)
  
  // Whether to delete expired items on get
  deleteExpiredOnGet?: boolean; // Default: true
}

Redis Cache Options

interface RedisCacheOptions {
  // Existing Redis client
  client?: IORedis;
  
  // Or connection options
  connection?: {
    host?: string; // Default: 'localhost'
    port?: number; // Default: 6379
    username?: string;
    password?: string;
    db?: number; // Default: 0
    tls?: boolean;
    url?: string;
  };
  
  // Error handling options
  handleErrors?: boolean; // Default: true
  maxReconnectAttempts?: number; // Default: 10
  reconnectStrategy?: 'linear' | 'exponential'; // Default: 'exponential'
}

Tiered Cache Options

interface TieredCacheOptions {
  // Array of cache providers in order of access (L1, L2, ...)
  providers: {
    provider: CacheProvider;
    options?: {
      writeOnHitFromLowerTier?: boolean; // Default: true
      ttl?: number;
    };
  }[];
  
  // Cache write strategy
  writeStrategy?: 'write-through' | 'write-behind' | 'write-around'; // Default: 'write-through'
  
  // Whether to propagate deletes to all tiers
  propagateDeletes?: boolean; // Default: true
  
  // Whether to propagate updates to all tiers
  propagateUpdates?: boolean; // Default: true
  
  // Maximum time in ms to wait for lower tier responses
  tierTimeout?: number; // Default: 5000 (5 seconds)
}

Invalidation Options

interface InvalidationOptions {
  // Invalidation method
  method: 'tag' | 'pattern' | 'version';
  
  // Tags for tag-based invalidation
  tags?: string[];
  
  // Pattern for pattern-based invalidation
  pattern?: string;
  
  // Namespace or version for version-based invalidation
  namespace?: string;
  version?: string | number;
  
  // Whether to cascade invalidation to related keys
  cascade?: boolean; // Default: false
  
  // Maximum number of keys to invalidate in one operation
  maxKeys?: number; // Default: 1000
}

Lock Options

interface LockOptions {
  // Whether to wait for the lock to be available
  wait?: boolean; // Default: false
  
  // Maximum time to wait for the lock in milliseconds
  waitTimeout?: number;
  
  // Time between retry attempts in milliseconds
  retryDelay?: number; // Default: 200
  
  // Maximum number of retry attempts
  maxRetries?: number; // Default: 10
  
  // Custom owner identifier for the lock
  owner?: string;
  
  // Additional metadata to store with the lock
  metadata?: Record<string, any>;
  
  // Whether to automatically extend the lock while held
  autoExtend?: boolean; // Default: false
  
  // How often to extend the lock in milliseconds
  autoExtendInterval?: number; // Default: 2/3 of TTL
  
  // Force acquire the lock, even if it's already held
  force?: boolean; // Default: false
}

License

MIT