Skip to content

Latest commit

 

History

History
94 lines (57 loc) · 2.2 KB

README.md

File metadata and controls

94 lines (57 loc) · 2.2 KB

NodeJS Cluster Abstraction Layer

This is an abstraction layer based on EventEmitter to provide a simple way of dealing with cluster support added on v0.6 of NodeJS.

Install

npm install rack

Usage

Don't forget to check the examples folder. Every available functionality is there.

var rack = require("rack").create();

rack.on("master-start", function () {
    console.log("master starting..");
});
rack.on("worker-start", function () {
    console.log("worker starting..");
});
rack.start();

Global Methods

.create()

Create a new Rack.

.cores()

Returns the number of detected CPU on your machine. This is used by Rack.start() if you do not specify the number of workers you want - it will use the number of CPU you have.

You can use this to start a multiple of workers based on CPU count.

var cluster = require("rack"),
    rack = cluster.create();

rack.start(cluster.cores() * 2);

Rack Events

master-start(rack)

This happens once you execute your script and will trigger only once for the master process.

worker-start(worker)

This happens for every worker that is started. Later on this can also happen when a worker dies and needs to start again.

worker-end(worker, rack)

This happens when a worker dies. You can start a new one by calling rack.worker().

worker-error()

This happens when a workers die too often.

message(message, worker, rack)

This happens when a worker sends a message to the master and vice-versa.

Rack Methods

.start(n)

Start a total of n workers.

.monitor(max, freq)

Start worker monitor and start a new worker whenever one dies. If max workers die during freq miliseconds, monitor will not restart the last one and will trigger worker-error event.

.worker()

This is the internal method to start a new worker. After calling .start(), if you want to start more workers just call this method.

.broadcast(msg)

Send a message from master to all workers. This will not work on workers.

.restart(interval [, cb ])

Kill all workers and start new ones with an interval (in ms) between them. If cb is specified, it will be called in the end of the restart process.