peavote

Easy-to-use polls website.

Preview after being deployed to AWS Elastic Beanstalk:

Elastic Beanstalk Dashboard after deployment:

Run locally:

Use Docker-Compose for running a local instance:

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

What's being used:

• Frontend
  • React
  • aura.css
• Backend
  • Express
  • Sequelize
  • Helmet
• Deployment
  • Docker
  • Docker Compose
  • Nginx
  • Travis CI
  • Docker Hub
  • AWS
    • AWS Elastic Beanstalk
    • AWS RDS (PostgreSQL)
    • AWS S3

Development Architecture:

Deployment Architecture:

Brief summary:

• Deployed on AWS Elastic Beanstalk with AWS S3 Buckets for storing Travis CI builds.
• Beanstalk is configured with a AWS RDS PostgreSQL instance running in Free Tier (db.t2.micro instance).

FAQ:

• Why not a Kubernetes deployment?
  • I actually wanted to deploy it through GCP using Kubernetes (GKE), however, my local Kubernetes environment had lots of issues (failed pod deployments, etc.) as I was using Kompose for converting my Docker-Compose deployment to a Kubernetes-based deployment. I eventually found out that AWS has Elastic Beanstalk, which is a much better solution for Docker-Compose based deployments, and hence, I deployed my application on Elastic Beanstalk.
• Why didn't you use PostgreSQL as a Docker image in your Elastic Beanstalk deployment?
  • I opted for AWS RDS as it easily manages backups, failovers and replications. It is also a bad practice for using Docker for storing persistent data.
• Is Kubernetes really required?
  • My answer would be no. Kubernetes is good, and I definitely would love to deploy my apps using it if I were to start from scratch, but honestly, AWS scales really well, so I'm not yet fully convinced for shifting to Kubernetes.