Implementation of a web-based data collection and aggregation infrastructure that utilizes secure multi-party computation techniques to allow individual contributors to submit their data without revealing it to the other participants.
It is expected that this application will operate on an Amazon Web Services EC2 instance running Amazon Linux under a security group that permits connections on port 80. The environment in which it runs is set up as follows.
- First, install Node.js, MongoDB, and necessary modules:
yum -y update
yum -y install nodejs npm --enablerepo=epel
echo "[MongoDB]
name=MongoDB Repository
baseurl=http://downloads-distro.mongodb.org/repo/redhat/os/x86_64
gpgcheck=0
enabled=1" | sudo tee -a /etc/yum.repos.d/mongodb.repo
yum install -y gcc-c++ mongodb-org-server mongodb-org-shell mongodb-org-tools`
- Navigate to the
server/directory, install the NPM dependencies, and install the global dependency:
npm install
npm install -g forever
- Next, set up the database file and start the MongoDB server:
mkdir -p /data/db
mongod
- Finally, retrieve the application files and in the directory "server/" run:
export NODE_ENV=production
authbind --deep forever -o log.txt -e error.txt start index.js
Instructions for setting up the environment on a Mac OS X or Linux local machine.
- Make sure to have Node.js and MongoDB installed.
- Navigate to the
server/directory and install the NPM dependencies:
npm install
- Now install the global dependency:
npm install -g forever
- Create the database file:
mkdir -p /data/db
- Start the MongoDB server:
mongod
-
Open a new terminal tab or window and navigate back to the
server/directory if you are not already there. -
Initialize JIFF:
git submodule init
git submodule update
cd jiff && npm install
- For testing, start the Node.js server with no environment variables:
node server/index.js
- For production, start the Node.js server with a production environment variable:
export NODE_ENV=production
node server/index.js
- Open up the browser and navigate to "localhost:8080"
This application can be used for a variety of deployments. Each deployment may have a different domain name and https certificate settings, as well as a different data format/layout.
server/config contains configuration files for each deployment specifying its https parameters and its data template. Data templates are json files typically located in client/app/data/, they are used to automatically render HTML UI and handle data aggregation.
The deployment is set to pacesetters by default, to change it, set a deployment environment variable"
export WEBMPC_DEPLOYMENT=deployment_nameIt is required that server/config/<deployment_name>.json is a valid configuration file. If the file is invalid, the server will fail on start.
Instructions on how to operate the web-mpc application. All steps below are performed in the browser.
- Navigate to
localhost:8080/create. - Click on Generate Session and save the two given files, one contains the session key and password which are needed for managing the session. The other contains a secret key needed to unmask the aggregate.
- Navigate to
localhost:8080/manage. - Input your session key and password.
- Generate participation links.
- Start the session.
- All participants will open a unique participation link, and proceed to fill out the information. Once completed, click Submit.
- Stop the session in
localhost:8080/manage. - Click the unmask link.
- Paste the session key and password in its designated fields.
- Click Browse and upload the private key file that was downloaded when generating the session key.
- Click Unmask Data and view the result.
Web-mpc is freely distributable under the terms of the MIT license. This release supports Handsontable's "Nested headers", a Pro feature. A valid license must be obtained when using this feature.