Public | Automated Build

Last pushed: a year ago
Short Description
Simple MEAN application with REST API for messages
Full Description


Simple MEAN application exposing a REST API for messages and determining whether they are palindromes. The API is REST-like in that messages are accessed and modified by using unique URLs that apply HTTP verbs: GET, POST, and DELETE.

The MEAN stack was chosen because of:

  • Node.js has good support for rapid development of HTTP servers with a REST-like API
  • Being a good platform for I/O-bound applications
  • Vast user community

It isn't a good platform for computationally intensive ones though.

Elixir is another great option for message based systems, but has a steeper learning curve.

Messages are stored in a Mongo database. Node.JS and Express acts as the server. Angular 2 serves the UI.

This app was generated with Angular CLI version 1.0.0-rc.2.


This app requires a Mongo database, or actually two: one for development, one for testing. The same database can theoretically be used for both development and testing, but that is discouraged.

Specify the development database via the environment variable MONGODB_URI.

Specify the testing database via the environment variable MONGODB_TEST_URI.

There are several free options for running a Mongo database:

Node.JS is also required.

Docker is required if the app will be run as a docker container.


Configurations are managed in the config/ folder. It is optional to change the content of these.

Build app locally

  1. Run npm install to build the project. The build artifacts will be stored in the dist/ directory.
  2. Make sure a Mongo database has been set up and specified via the appropriate environment variable, see Configuration section.
  3. Run npm start to start the server.
  4. Navigate to http://localhost:8080/.

Update API documentation locally

  1. Run npm run apidoc to generate the API documentation. Help will be generated in the dist/apidoc directory.
  2. Navigate to http://localhost:8080/apidoc.

Running functional tests

Run npm test to execute functional tests via Mocha.


The app is available on Docker Hub.

The source code also contains a Dockerfile for building an image locally. There is also a .dockerignore file.

Pull image from Docker Hub

docker pull karljohangrahn/rotavator

Build image

docker build -t <your username>/rotavator .

Run image

  1. Make sure a Mongo database has been set up and specified via the appropriate environment variable, see Configuration section.
  2. docker run -d -e MONGODB_URI="mongodb://<MONGODB_URI>" -p 8080:8080 <IMAGE ID>
  3. Navigate to http://localhost:8080/.

Deploy to AWS

  1. Create an AWS account and put the AWS_ID, AWS_SECRET, and AWS_REGION in corresponding environment variables.
  2. Generate an SSH RSA (not DSA) key with ssh-keygen if there is not already one.
  3. Create a VM:

    $ node_modules/.bin/awsbox create -n myvm

  4. Deploy code:

    $ git push myvm HEAD:master

  5. Visit http://<my ip address>.

Deploy to Google Cloud

  1. Use the Cloud Platform Console to create a new Cloud Platform project, create an App Engine application, and enable billing.
  2. Download and install the Google Cloud SDK and then initialize the gcloud tool.
  3. Modify app.yaml and add URI for Mongo database.
  4. Deploy by running the following command from the GIT directory:

    gcloud app deploy

  5. View the app at

Sequence diagrams

Show the list of messages posted by the users

Post a new message

Select a given message to see extra details



Docker Pull Command
Source Repository