Streaming Cache

Speed up your services.

Cache, queue and distribute streams immediately. Streams can be replayed immediately, even if the source is not finished.

Uses a fixed size LRU-cache in the background.

Usefull for caching (slow) streaming connections, such as S3 requests or complex database queries.

If there are for example 3 requests for a certain file in close succession, then we can pipe the result for the first request into the cache. The 2 other requests will receive a stream which will start even before the first one is finished.

Performance

Serving from this cache is extremely fast. On my local machine I get 2.5GB per second for a single process on localhost using AB. (4th gen i7).

Installation

npm i streaming-cache --save

Quick example

varCache=require('../index.js');varcache=newCache();varfs=require('fs');varinputStream=fs.createReadStream('readme.md');varoutputStream=fs.createWriteStream('test2.txt');inputStream.pipe(cache.set('myKey'));setTimeout(function(){outputStream.write('written from cache\n\n');cache.get('myKey').pipe(outputStream);},200);

Options

For a list of options see: https://www.npmjs.com/package/lru-cache

API

set(key)

returns a Duplex stream

fileStream.pipe(cache.set('key')).pipe(res);

get(key) => ReadableStream

varcached=cache.get('key');if(cached){cached.pipe(res);}

setData(key, data) => WriteableStream

A set data synchronously to stream at a later moment

getData

Get data with a callback.

cache.getData('key',function(err,data){if(err){//handle error}// do something with data}));

setMetadata(key, data)

Set metadata for a stream to be used later.

getMetadata(key, data)

Get metadata

exists(key)

returns true or false if a key exists.