This repo contains Web-based simple frontend and backend for TrojanChat Application, which is the final project of EE547 at USC.

• Frontends: One Windows Application built with WPF using .NET6 (Not in this repo), one Web Application built with HTML, CSS, Boostrap, and AJAX.
• Backends: Both backend 1 and backend 2 are built with NodeJS using express.
• RESTful API: Interaction between Frontend and Backend 1.
• GraphQL API: Interaction between Backend 1 and Backend 2.
• Kafka Pipeline: Backend 1 as Topic Producer; Backend 2 as Topic Consumer.
• MongoDB: Text messages and User information storage.
• AWS S3: Image and Video storage.
• AWS ECS: Project deployment (backends and web-frontend).

Cloud Prerequisite

• A Standard S3 bucket will suffice for storage.
• A single t2.small EC2 server (Linux, x86 64-bit, 1 vCPU, and 2 GiB Memory), which is not in free-tier and costs 0.023 USD per Hour for the time I do.

Note: 1 GiB memory does not suffice, because bitnami/zookeeper:latest needs more than 1 GiB.

Env. Prerequisite

Once you get this repo to AWS already, you may need to:

• Install Docker

sudo yum update -y
# install docker
sudo amazon-linux-extras install docker
# launch service
sudo service docker start
# grant permission to user
sudo usermod -a -G docker ec2-user

• Install Docker Compose

# install latest version
sudo curl -L https://github.com/docker/compose/releases/latest/download/docker-compose-$(uname -s)-$(uname -m) -o /usr/local/bin/docker-compose
# grant permission to docker-compose
sudo chmod +x /usr/local/bin/docker-compose

Change Codes

• Web Frontend

For both frontend -> index.html and frontend -> chat_app.html:

Change the line of const serverIP = '54.187.66.112'; to the IP of the server you are using.

• AWS S3

For producer -> .env:

Change these fields: AWS_BUCKET_NAME, AWS_BUCKET_REGION, AWS_ARN, AWS_ACCESS_KEY, AWS_SECRET_KEY.

Where the AWS_ACCESS_KEY and AWS_SECRET_KEY are the key of the user you created for accessing AWS S3.

Launch Project

Go to the root of project, where docker-compose.yml exists.

• Launch Docker Containers

docker-compose up -d

• Install Packages

# enter backend1
docker exec -it backend1 bash
# install packages
npm i

• Launch producer/backend1

npm run start:producer
# or
node producer/index.js

• Enter and Launch consumer/backend2

npm run start:consumer
# or
node consumer/index.js

You are ready to go now!

Go to browser: http://ip_of_your_server:3000/

If this cannot access, you may need to open the port for your server.

AWS -> EC2 -> Network & Security -> Security Groups

Choose the one that your server are using, then give it Custom TCP with port 3000 (Source could be either 0.0.0.0/0 or specific to your local IP).

For running locally, AWS S3 are still needs to be setup as mentioned above. But the IP in frontend files could be just localhost.

If on Windows, could use Docker Desktop for launching containers.

• Directories

consumer: Consumer code (Backend 2), including GraphQL API implementaion, Kafka Consumer, MongoDB Client.
demo: Images for demonstration.
frontend: Web-based Frontend code.
producer: Producer code (Backend 1), including RESFful API implementation, AWS S3 client, Kafka Producer, GraphQL client.
requests: VScode RESTful API debugging tool.

• Files

dirname.js: Helper file for get current path location.
docker-compose.yml: Docker-Compose file for launching containers, including zookeeper, Kafka, Backend 1, Backend 2, and MongoDB.
graphql.txt: A record of a few GraphQL requests used for debugging when developing.
messageType.js: The message format definition for streaming through Kafka Pipeline.
package.json: Packages Dependencies.