Seneca-balance-client-ng NPM

A Seneca.js transport plugin that provides various client-side load balancing strategies, and enables dynamic reconfiguration of client message routing.

seneca-balance-client

Description

This module is a plugin for the Seneca framework. It provides a transport client that load balances outbound messages on a per-pattern basis.

If you're using this module, and need help, you can:

Post a github issue,
Tweet to @senecajs,
Ask on the Gitter.

If you are new to Seneca in general, please take a look at Senecajs.org. We have everything from tutorials to sample apps to help get you up and running quickly.

Seneca compatibility

Supports Seneca versions 3.x and above.

Install

npm install seneca-balance-client

And in your code:

require('seneca')()
  .use('balance-client', { ... options ... })

Quick Example

server.js

require('seneca')()

  .listen( {port: function () { return process.argv[2] }} )

  .add('a:1', function (msg, done) {
    done( null, {a: 1, x: msg.x} )
  })

// run twice:
// $ node server.js 47000 --seneca.log=type:act
// $ node server.js 47001 --seneca.log=type:act

client.js

require('seneca')()
  .use('balance-client')

  .client( {type: 'balance'} )
  .client( {port: 47000} )
  .client( {port: 47001} )

  .ready( function () {

    for ( var i = 0; i < 4; i++ ) {
      this.act( 'a:1,x:1', console.log )
    }

  })


// $ node client.js --seneca.log=type:act

The client will balance requests over both servers using round-robin. As there is no pin in the .client configuration, this will apply to all non-local actions. Add a pin to restrict the action patterns to which this applies - make sure to use the same pin on both client and server to avoid ambiguity.

Usage

The plugin provides two balancing models:

consume: messages are sent to individual targets, using a round-robin approach
observe: messages are duplicated and sent to all targets

You specify the model using the plugin option model:

var Seneca = require('seneca')

var s0 = Seneca({tag: 's0'})
  .listen(44440)
  .add('a:1', function (msg, done) {
    console.log('s0;x='+msg.x);
    done()
  })

var s1 = Seneca({tag: 's1'})
  .listen(44441)
  .add('a:1', function (msg, done) {
    console.log('s1;x='+msg.x);
    done()
  })

var c0 = Seneca({tag: 'c0'})
  .use('..')
  .client({ type: 'balance', pin: 'a:1', model: 'observe' })
  .client({ port: 44440, pin: 'a:1' })
  .client({ port: 44441, pin: 'a:1' })


s0.ready( s1.ready.bind(s1, c0.ready.bind(c0, function () {
  c0.act('a:1,x:y')

  // wait a little bit to avoid shutting down in mid flow
  setTimeout(
    s0.close.bind( s0, s1.close.bind(s1, c0.close.bind(c0))), 111 )
})))

You can also provide your own balancing model by providing a function with signature (seneca, msg, targetstate, done) as the value of the model setting:

...
    .client({
      type: 'balance',
      pin: 'a:1',
      model: function (seneca, msg, targetstate, done) {
        if (0 === targetstate.targets.length) {
          return done( new Error('No targets') )
        }

        // select a random target
        var index = Math.floor(Math.random() * targetstate.targets.length)
        targetstate.targets[index].action.call( seneca, msg, done)
      }
    })
...

The targetstate object provides you with the list of currently available targets. Review the internal implementations of the observeModel and the consumeModel in balance-client.js for a starting point to write your own model.

Contributing

The Senecajs org encourages open participation. If you feel you can help in any way, be it with documentation, examples, extra testing, or new features please get in touch.