Node based Data Visualization Server built on phantom-lambda-template

The bare minimum for a phantomjs app running on Amazon Lambda. Based off of node-lambda-template

• Keep in mind that aws-lambda runs on Aws Linux, you will have to make sure your package is compiled for Aws-Linux

The phantomjs program has been build to for Linux, https://bitbucket.org/ariya/phantomjs/downloads/phantomjs-1.9.8-linux-x86_64.tar.bz2 If you are going to run this locally you will have to swap out phantomjs with the assembly built for your OS.

It uses node-lambda under the hood to locally run and also deploy your node.js Amazon Lambda application.

./node_modules/.bin/node-lambda run

Usage

There are 3 available commands, make sure to use the appropriate phantomjs for your OS.

./node_modules/.bin/node-lambda setup
./node_modules/.bin/node-lambda run
./node_modules/.bin/node-lambda deploy

Install

git clone https://github.com/justengland/phantom-lambda-template.git
cd node-lambda-template
npm install

Package

zip -r phantom.zip . --exclude=.DS_Store --exclude=.git --exclude=node_modules --exclude=.idea --exclude=.c9

Install Phantomjs

npm install phantomjs --phantomjs_cdnurl=http://cnpmjs.org/downloads

Ideas implemented in the code

1. Install phantom locally using NPM - this will install the phantomjs version for your OS into node_modules
2. Package for Lambda without the node_modules, then the node script will read from the disk, if node_modules does not exists then read it from the root of the package which will be setup for lambda

Needs to improve

1. Need to try with a bigger dataset.
2. Need to try to have Phantomjs running as a background service or server. Maybe not using Lambda and using ec2 instead would be a better approach for this?
3. Using the cluster api of node and making sure clustering is taking place irrespective of whether its on AWS Lambda or AWS EC2.

Here's the output of "time node run.js":

real	0m11.392s
user	0m0.758s
sys	    0m0.193s