I've been exploring and using different cloud services for a few years now. Starting with the Google Cloud Platform then working my way into being a daily user of Microsoft Azure it was AWS Step Functions that truly converted me. Utilizing the cloud can be a cheap and scalable way to manage your application. To demonstrate some of the features of AWS in a fullstack sense, I built a boilerplate React/Typescript application. I'll go through and explain each stage and the tools I used. I hope to show others what I've learned and hopefully get some feedback to continue to expand on my understanding of these tools. Keep in mind I'm still wrapping up the build, but I wanted to go ahead and document the process.

Stage 1: Frontend

The goal of the application, as stated above, is to demonstrate the front-to-back flow of building a full-stack serverless application. So rather than build out complex features, I stuck to the basics. A login page, a registration page, and a simple page that demonstrates some selected details about the user, their favorite color, a favorite song, etc. I built this using React.js, Typescript, TailwindCSS, and AWS Amplify for quick and easy hosting.

Stage 2: API

For the API, I'm using AWS API Gateway using Lambda functions. I've enjoyed using the gateway with Lambda. Lambda functions are easily scalable and very cheap. For this project, there are a few lambda functions responsible for authenticating a user's email and confirming the hash of the email + password + salt for login using SHA256.

Stage 3: Backend

As for the back, I'm using DynamoDB for all things user related. I thought NoSQL made sense for this. As for all of the details about the application: colors, songs, etc. I've organized these in a PostgreSQL relational database service. This way when you're selecting a color or a song, we can use the ID of that color in the corresponding table. I also used this as a way to get more comfortable setting up and using RDS being that I've relied heavily on NoSQL for most of my projects.

Stage 4: Documentation

For documentation I'm using stackedit.io to write out the markdown, I'm using my website to host the content, and cloudcraft.co to design visuals of the architecture.

Stage 5: Security

The main points of the security in this application would be making sure that there are no sensitive details exposed in the code and controlling access to the API Gateway. For this, we make sure to store sensitive information in the environment variables of our lambda functions and the Amplify application itself. The other piece of this is securing the REST API. For this, I'm using standard IAM roles and policies to control, edit and access the gateway. I'm also considering the use of Lambda authorizers to better control who is allowed in and out.

Stage 6: DevOps

This step is pretty straightforward being that, I point Amplify to the GitHub branch, mine being "Main", that the application lives in. Every time I push a new version into that repository, the application's build is automatically triggered again.

Final Thoughts:

The process of building out this application is eye-opening mainly because of the ease of use. I recall, the setup and management of Node.js servers and getting them deployed, and thinking about scalability and it can quickly feel like a headache. I've found that using AWS has taken a lot of the headache out of dealing with the backend. It's also quick and easy. I'd love to hear your thoughts on how I've set this up. Agreements and disagreements are welcome. Thanks for reading!