Simple Message Store App

by Ram Darbha

Architecture

Node.js and Express.js web application frameworks
Mongodb for NoSQL storage
Docker, and Docker Compose for portable development and deployment
Mocha Javascript test framework and Chai assertion library for testing

This is an implementation of a simple headless web server that presents a RESTful API for operations to add, delete, retrieve and list text messages. The node.js/express.js frameworks implement the web service while a mongodb database provides the persistent store.

API

Operation            HTTP Method   URI                                 Response

Get message 23       GET           http://localhost:8081/msg/23        200; 404 if not found
                                                                       Body:
                                                                       {
                                                                           id: "23",
                                                                           msg: "asd fadf adsf adsf asd f",
                                                                           palindrome: "no"
                                                                       }
                                                                    
Delete msg 23        DELETE        http://localhost:8081/msg/23        200; 404 if not found

Delete all msgs      DELETE        http://localhost:8081/msg/all       200;

List all messages    GET           http://localhost:8081/msg           200;
                                                                       Body:
                                                                       [
                                                                           {
                                                                               id: "23",
                                                                               msg: "asd fadf adsf adsf asd f",
                                                                           },
                                                                           {
                                                                               id: "25",
                                                                               msg: " ghjkjghkhjgkjkjh",
                                                                           }
                                                                       ]

Add message          POST          http://localhost:8081/msg           200 if ok
                                   Body:                               Body:
                                   abc cba                             {
                                                                           id: "23",
                                                                           msg: "abc cba"
                                                                       }

Object Model

The object model in this app is very simple: a single collection called "messages" with 2 fields:

message: a text field to store the user-supplied message
_id: a UUID generated by MongoDB

How to Setup

Install Docker v1.11.0+
Install docker-compose v1.7.0+
Clone this repo: git clone https://github.com/darbha-ram/msgstoreapp.git msa_dir

How to Build and Deploy

docker-compose uses the specification in docker-compose.yaml to build the web image, deploy the db container, then deploy the web container linking it to the db.

Change directory: cd msa_dir
Build and deploy the web app and database containers: docker-compose up

How to Build and Deploy Manually

If you choose, these are the steps to deploy the containers directly, without using docker-compose.

Change directory: cd msa_dir
Build web app: docker build --tag "msa" .
Start Mongo db: docker run --name db -d -p 27017:27017 mongo:3.0.11
Start web app: docker run --name web -d -p 8081:8081 --link db msa

Testing

Change directory: cd msa_dir
Run test: npm test

Notes

Implementation is based on public node:argon and mongo:3.0.11 docker images.
Public images will be pulled down by docker the first time they are needed for build or deployment. Thereafter they will be available locally, can be viewed using docker images.
The model and controller (API) are very simple and in a single file index.js.
Server does not track users, so any HTTP request can view or operate on any messages.
These Node.js extensions are needed to build & deploy the web app: express, mongodb, body-parser.
In addition, these Node.js extensions needed to develop & test: mocha, chai, request.
All these extensions are captured in package.json and installed by npm install when the web container starts up.

Future Extensions

API and data model can be extended to include the notion of users e.g. /msg to /user/msg, and a storing a user field in each message document.
To evolve to a more complex model, Mongoose ODM can be used to model and validate the schemas.

darbha-ram / msa