DADI Cache

A caching layer supporting Redis and filesystem caching, used in core DADI products

• Overview
• Install
• Usage
  • Create Cache instance
  • Add an item to the cache
  • Get an item from the cache
  • Example real world usage
• Configuration
  • General Options
  • Default Options
  • Filesystem Caching
  • Redis Caching
  • Redis Cluster
• Cache Fallback

Overview

Removing the complexity involved in setting up two separate cache handlers for every project, DADI Cache can use either the filesystem or a Redis instance to store and retrieve content.

Install

npm install @dadi/cache --save

Usage

Create Cache instance

// require the module
const Cache = require('@dadi/cache')

// setup the options for caching
// defaults if nothing specified: 
// {
//   directory: {
//     enabled: true,
//     path: './cache'
//   },
//   redis: {
//     enabled: false
//   }
// }
const options = {
  "ttl": 3600,
  "directory": {
    "enabled": false,
    "path": "./cache/"
  },
  "redis": {
    "enabled": true,
    "host": "127.0.0.1",
    "port": 6379
  }
}

const cache = new Cache(options)

Add an item to the cache

set(key, data)

Returns a Promise that returns an empty String if successful, otherwise an Error.

The data argument can be a String, Buffer or Stream.

let key = 'test-cached-item'
let data = 'test data'

cache.set(key, data).then(() => {
  // do something
}).catch(err => {
  // Error
})

Get an item from the cache

get(key)

Returns a Promise that returns a Stream of the cached data if the key exists or an Error if it does not exist. The error message returned is "The specified key does not exist".

let key = 'test-cached-item'

cache.get(key).then(stream => {
  // do something with the stream
}).catch(err => {
  // err === "The specified key does not exist"
})

Example real world usage

const express = require('express')
const app = express()
const dadiCache = require('@dadi/cache')
const cache = new dadiCache()

app.get(function (req, res) {
  let key = req.url

  cache.get(key).then(stream => {
    // cached data found for req.url
    res.setHeader('X-Cache', 'HIT')
    stream.pipe(res)
  }).catch(err => {
    // cached data not found for req.url
    let content = fetchContent()

    // cache the content
    cache.set(key, content).then(() => {
      res.setHeader('X-Cache', 'MISS')
      res.end(content)
    })
  })
})

Configuraton

General options

Property|Description|Default|Example --------|-----------|-------|------- ttl|The time, in seconds, after which cached data is considered stale|true|3600

Default options

A DADI Cache instance can be created with no options, in which case the following options will be used:

{
  "directory": {
    "enabled": true,
    "path": "./cache"
  },
  "redis": {
    "enabled": false
  }
}

Filesystem caching

Property|Description|Default|Example --------|-----------|-------|------- enabled|If true, caching is enabled using the following settings|true| path|The absolute or relative path to the directory for cache files|"./cache"|"/tmp/dadi-cache/" extension| (optional) The extension to use for cache files| none | "json" directoryChunkSize| (optional) If set, cache files are stored in a series of subdirectories based on the cache key| 0 | 5 autoFlush | If true, DADI Cache will clear cache files that are older than the specified TTL setting, at the interval specified by autoFlushInterval | false | true autoFlushInterval | The period of time between clearing cache files (in seconds) | 300 | 1800

Redis caching

A set of options for both file and Redis caching must be provided if you intend to use Redis as the cache store. This allows DADI Cache to fallback to file caching in the event of a Redis connection failure.

Property|Description|Default|Example --------|-----------|-------|------- enabled|If true, caching is enabled using the following settings|false|true host|The hostname or IP address of the Redis server |"127.0.0.1"|"drum.redistogo.com" port|The port of the Redis server|6379 |9092

{
  "directory": {
    "enabled": true,
    "path": "./cache"
  },
  "redis": {
    "enabled": true,
    "host": "127.0.0.1",
    "port": 6379
  }
}

Redis Cluster

Property|Description|Default|Example --------|-----------|-------|------- enabled|If true, caching is enabled using the following settings|false|true cluster|If true, DADI Cache will connect caching is enabled using the following settings|false|true scaleReads|Specify where to send queries, to the masters, slaves, or a combination. See Read-Write Splitting |"master"|"all" hosts|When cluster: true, Cache uses this array of hosts to connect. Each array item must contain a host and port.|| [{"host":"127.0.0.1", "port": 6379}, {"host":"127.0.0.1", "port": 6380}]

To connect to a Redis cluster an array of hosts must be specified, rather than a single host and port.

The array does not need to contain all your cluster nodes, but a few so that if one is unreachable the next one will be tried. DADI Cache will discover other nodes automatically when at least one node is connnected.

{
  "directory": {
    "enabled": true,
    "path": "./cache"
  },
  "redis": {
    "enabled": true,
    "cluster": true,
    "scaleReads": "all",
    "hosts": [
      {
        "host": "127.0.0.1",
        "port": 6379
      },
      {
        "host": "127.0.0.1",
        "port": 6383
      }
    ]
  }
}

Read-Write Splitting

A typical Redis cluster contains three or more masters and several slaves for each master. It's possible to scale out Redis cluster by sending read queries to slaves and write queries to masters by setting the scaleReads option.

scaleReads is "master" by default, which means no queries will be sent to slaves. The other available options:

• "all": Send write queries to masters and read queries to masters or slaves randomly.
• "slave": Send write queries to masters and read queries to slaves.

For example, with scaleReads: "slave":

cache.set('foo', 'bar') // This query will be sent to one of the masters.
cache.get('foo', (err, res) => {
  // This query will be sent to one of the slaves.
})

Note: In the code snippet above, the result may not be equal to "bar" because of the lag of replication between the master and slaves.

Cache Fallback

In the case of a Redis connection failure, DADI Cache will attempt to reconnect four times before switching to file caching. After a configurable period (default 5 minutes), an attempt will be made to reconnect to Redis and if successful DADI Cache will resume using Redis as the cache store.

Roadmap

• Add LATENCY HISTORY to assist in determining Redis performance
• Add FLUSH
• Add authentication options for Redis