Using deployd with redis

server.js

var deployd = require('deployd');

var redisAdapter = require('socket.io-redis');
var redis = require('redis');

// need two client pairs because one is used for socket.io and the other for sessions
// might not be necessary anymore, see https://github.com/socketio/socket.io-redis/issues/90
var pub = redis.createClient(process.env.REDIS_URL);
var sub = redis.createClient(process.env.REDIS_URL, { 'return_buffers': true });

var pub2 = redis.createClient(process.env.REDIS_URL);
var sub2 = redis.createClient(process.env.REDIS_URL);

var server = deployd({
  port: process.env.PORT || 1337,
  env: process.env.ENV || 'production',
  origins: ['http://localhost:5000', ... ], // optional
  allowedResponseHeaders: ['ETag'], // optional
  allowedRequestHeaders: ['If-None-Match', 'X-App-Id'], // optional
  db: {
    connectionString: process.env.MONGO_URI,
  },
  sessions: {
    pubClient: pub2,
    subClient: sub2,
  },
  socketIo: {
    options: {
      //transports: ['websocket'], // optional
    },
    adapter: redisAdapter({ pubClient: pub, subClient: sub }),
  },
});

// start the server
server.listen();

// debug
server.on('listening', function() {
  console.log("Server is listening on port: " + (process.env.PORT || 1337));
});

// Deployd requires this
server.on('error', function(err) {
  console.error(err);
  process.nextTick(function() { // Give the server a chance to return an error
    process.exit();
  });
});