DHC

Note:

This is a mirror of the original repo since I can't make it public

Setup

If you have not already, clone the repo to get the source code.

git clone https://github.com/SCCapstone/DirtyHistoryCrawler.git

Enter the dockerized_webapp directory.

cd dockerized_webapp

Dependencies

This app uses:

Node.js
Polymer.js
MongoDB
Neo4j

To make our lives and the lives of those who wish to run this software easier, we have chosen to deploy our app inside Docker containers, but, it is still best to run in a linux environment.

Install Docker:

Building

You need a configuration file which defines the port the webapp will be served on. We have provided an example file to model what this should look like.

Before deployment, be sure to edit the .env.example and rename as .env. The naming convention is very important. You MUST name this file .env or docker-compose will not recognize it.

Data Acquisition:

If you want to run the worldcat miner, you need a wskey from worldcat. Once you have one, set the WSKEY environment variable in your .env file.

Production Versus Development Builds:

In the dockerized_webapp directory there are two different .yml files to configure docker-compose builds. docker-compose.yml is the configuration file for production builds and is the default configuration file. The main difference between this and the docker-compose.dev.yml file is that the former has more memory allocated to neo4j and is configured to restart containers if they exit with a non-zero code. Also, the dev configuration has opened ports for the neo4j web browser, which is convenient for doing development, but insecure for public facing production deployments.

To build the production version of the app:

$ docker-compose build

To build the development version of the app:

$ docker-compose -f docker-compose.dev.yml build

Deploying

for production:

$ docker-compose up -d

for development:

$ docker-compose -f docker-compose.dev.yml up -d

The app will be deployed on localhost:$SERVER_PORT.

Notes:

If you want, you can skip the build command and just run up. Compose will automatically run the build step. However, sometimes you may want to separate them out to tweak the build process; e.g. running docker-compose build --no-cache to ensure that the build is completely fresh when your source code is altered.
You can check on the status of your containers with docker logs <container_id>. The id can be found by running docker ps.
If you specified the docker-compose.dev.yml config file to run a development build, you must also sepcify this file if you want to stop the deployment: docker-compose -f docker-compose.dev.yml down.

Testing

For Behvaioral Tests:

Tests must be ran locally.

cd into dockerized_webapp\public\
Ensure npm and bower are installed globally. Then, install polymer and the needed bower components using:

$ npm install -g polymer-cli
$ bower install

Finally, run the test:
$ polymer test

polymer test only logs if a test fails by default, to show a log for each test ran use:

$ polymer test --expanded

You may need to use sudo before each command if you run into permission issues. If you run into any issues with the polymer-cli email me at vmcquinn@email.sc.edu

Currently, we don't have a browser set up with docker to run the tests with selenium hence testing locally.

For Unit Tests:
Install mocha globally with:
$ npm -g mocha Then run:
$ mocha

Potential Problems

If you have MongoDB installed locally on your machine, it may be running in the background. If so, it is likely running on the port that the dockerized instance of MongoDB from this app is intended to run on. If you run into this problem, there are two quick fixes:

You can stop your local mongo instance. I do this on ubuntu with systemctl stop mongod.
You can go into the docker-compose.yml file and change the port mapping for mongo to a different port.

jdrain/DHC-Mirror