Twitter-App

Team 1

Name	Email Address
Rishi Jatania	jatania.r@northeastern.edu
Karun Kesavadas	kesavadas.k@northeastern.edu

Introduction:

Twitter-App is an the modern microblogging and social networking service on which users post and interact with messages known as "tweets". Registered users can post, like and comment on tweets, follow other users for daily updates. The app also blocks offensive tweets and comments by users using RapidAPI module called Purgomalum. Other features include Abusive tweet and Bug reporting in Slack Incoming Webhooks.

Deployment scripts recreate this scalable microservices application and deploy it on AWS CLoud Infrastructure.

Scope:

The deployment scripts would create 2 worker nodes within AWS EKS cluster automated using Terraform scripts. All routing is managed internally by Service Names, so no IP's needs to be edited.

Features

Application Features – Twitter

1. Token based authentication for user Sign in and Sign up
2. Web application where user can tweet and see the recent tweet from other users who they follow.
3. Tweets include photo, username, tweet body, comments and likes on a particular post.
4. User can comment and like the tweet and view tweets posted by others
5. Ability to delete offensive tweets before any user can see them
6. All the tweets are store to the backend database, which will help the user find their past tweets.
7. Ability to deploy front-end as a PWA so that it can run on devices
8. Redis queue for caching requests.
9. React and material UI for frontend development

Application Deployment

1. Terraform to provision infrastructure
2. Kubernetes cluster on AWS
3. Monitoring using Prometheus
4. Continuous Integration to Docker Hub (A new image is created and deployed to docker hub) using Github Actions
5. Deploy app on cloud with a single command (automation)
6. Grafana Dashboard for visualization

Infrastructure:

The project is deployed on Kubernetes cluster and has the following components:

Components:

- Frontend:
	1. Frontend deployment
	2. Frontend Load balancer Service
- Middleware:
	1. Middleware deployment
	2. Middleware Load balancer Service
- Redis-Queue:
	1. Queue deployment
	2. Queue Load balancer Service
- Backend:
	1. Backend Auth/Profile Deployment
	2. Backend Auth/Profile Load balancer Service
	3. Backend Tweet Deployment
	4. Backend Tweet Load balancer Service
- MongoDB: - (This is a hosted cluster on MongoDB Atlas):
	1. MongDB Atlas cluster
- Prometheus:
	1. Prometheus deployment
	2. Prometheus Load balancer Service
- Grafana:
	1. Grafana dashboard for visualization
- Metrics-Server
	1. Source of Container resource metrics for AutoScaling

PreReq tools that you need

1. aws-cli
2. kubectl
3. git
4. terraform
5. aws-iam-authenticator

Dependancies

We need Terraform, Docker for running the deployment. We need S3 Bucket Name for hosting images, Slack WebHooks, RapidAPI credentials, RAPID API KEY for Purgomalum, DockerHub Username and Password and AWS Account Access Key and Secret of the Account that will run the EKS stack and also has the S3 Bucket. We also need the MongoDB Atlas Cluster URL.

Continuous Integration Setup

In-order to setup continuous integration in github using github actions please configure your Docker Username(DOCKER_USER) and Password(DOCKER_PASSWORD) by following these steps

Initial set up

The initial steps with screenshots can be found in the Project Deployment Steps PDF.

OR

Clone the project, and open the Terminal(Linux/Mac) or Git bash(Windows) into the infrastructure directory of the project and run the command

./deployment.sh

The scripts deploys the infrastructure using terraform scripts and ask for the details required to populate the secrets in the following order:

1. MongoDB Cluster URI
2. MongoDB Name
3. Slack Webhook URI
4. JWT Secret (Can be any random string)
5. Rapid API Key
6. AWS Access Key
7. AWS Secret Access Key
8. S3 Bucket Name

Once secrets have been generated the script execution continues and deploys K8 cluster services.

Once the cluster is created check the services with:

kubectl get svc

To Access the Front-end use the link for the LoadBalancer of twitter-front end.

To Open Grafana Dashboard use the following command and follow the url with the port mentioned.

kubectl get svc -n monitoring

PWA Installation steps:

You will find a '+' symbol to right of the URL. Cick on it and install the app on to your device. If you don’t find a '+' to the right of the URL bar after going in to the deployed website then follow these steps.

1. Open Chrome browser and type the following in the url (chrome://flags/#unsafely-treat-insecure-origin-as-secure)
2. Enable the "Insecure origins treated as secure" by adding the deployed application's URL in the input field.
3. Now reload browser and you will see '+' symbol to right of the URL. Cick on it and install the app on to your device.

Delete Cluster:

To delete the cluster run the destroy.sh script found in the infrastructure directory.

./destroy.sh

Docker Images:

• rishijatania/twitter-backend
• rishijatania/twitter-frontend
• rishijatania/twitter-middleware
• rishijatania/twitter-tweet