← Back to Blog

Why I Moved My Side Project Off AWS

Published May 4, 2026 · 6 min read · Galaxy Cloud Solutions

I want to be clear that AWS is genuinely impressive. The scale of what they have built, the breadth of services, the reliability at enterprise level — it is all real. I am not going to pretend otherwise just because I am going to tell you about leaving it.

But I had a side project on AWS and I moved it off, and the reason was simple and a little embarrassing: I was spending more time reading AWS documentation and more money on services I barely used than was justified by anything the project was doing.

⚡ VPS from $5/mo — Use code LAUNCH2026 for 50% off

The Bill That Made Me Think

The project was a small web app. A few hundred users, maybe 50 active ones on any given day. Nothing demanding. I was running it on an EC2 t3.small with an RDS Postgres instance and S3 for file storage.

The monthly bill was $47. For a side project with no revenue. Every month I would open the AWS bill, feel a mild sense of dread, and think "I really should do something about this."

The thing that finally tipped me over was when I accidentally left a NAT Gateway running for a week. That one mistake added $18 to a single bill. I did not fully understand what a NAT Gateway was, I had created it following a tutorial, and I had no idea it was sitting there accumulating charges. That is not how running infrastructure should feel.

The AWS Complexity Tax

AWS has a service for everything. That is both its greatest strength and its biggest problem for small projects. There is almost no simple way to do anything — there is the AWS way, which involves IAM roles, VPCs, security groups, availability zones, and a mental model of how all these pieces fit together that takes months to build properly.

I spent more time managing AWS than building the project. Learning what a VPC was, figuring out why my EC2 instance could not reach the internet, debugging IAM permission errors that made no sense, understanding why my S3 bucket policy was rejecting requests. All of that time was not making the project better. It was managing the platform.

What the Migration Looked Like

I moved to a $10 VPS. The migration took one afternoon. Nginx handles the web traffic, Postgres runs on the same machine, and the uploaded files live on the VPS disk with a daily rsync backup offsite. The whole setup is simpler and I understand every piece of it.

The monthly bill is $10. I went from $47 to $10 and the project runs identically. There is no meaningful difference in performance for an app with 50 daily active users.

When AWS Is the Right Answer

I want to be fair here because this is not an anti-AWS piece. AWS is the right answer when you need global scale, compliance requirements that mandate enterprise infrastructure, tight integration with other AWS services, or a team large enough to manage the complexity. These are legitimate needs and AWS handles them better than anything else.

For a solo developer with a side project or an early-stage startup trying to keep costs down, AWS is frequently the wrong tool. The complexity is real, the bill management is stressful, and the cognitive overhead of the platform competes with the cognitive overhead of building your actual product.

What I Would Do Differently

Start simple. A VPS is simpler than AWS in almost every way. You SSH in, you install what you need, you run your app. No IAM roles, no VPCs, no availability zones, no NAT gateways accidentally left running. When your project actually needs AWS-scale infrastructure, you will know — and you will have the revenue to justify it. Until then, the $47/month you save is real money.