Posted on my personal blog · Cross-posted from AWS Builder Center
A few days ago, I published something a little different on the AWS Builder Center — not a how-to guide, not a step-by-step tutorial, but more of a reflection on a service we all take for granted.
The article is called:
"11 Things About S3 Most AWS Developers Never Stop to Question"
You can read the full version here 👉 AWS Builder Center
Here's the honest truth: I'd been using S3 for years before I stopped and actually thought about what it is.
Not how to use it — I knew that. But what it actually is underneath. What makes it tick. Why it's designed the way it is. And what patterns you unlock once you stop treating it like a USB drive in the cloud.
So I wrote the thing I wish someone had handed me three years ago.
The full article goes deeper, but here's the core of what I explored:
S3 has no folders. That "folder" you created? It's a lie — a helpful one. S3 is a flat key-value store, and once you understand that, you stop fighting the design and start using it properly.
The durability number is genuinely wild. Eleven nines — 99.999999999%. If you stored 10 million objects, you'd expect to lose one every 10,000 years. And you get this for about $0.023 per GB per month. The maths of what that would cost to replicate on-premises should make any infrastructure person uncomfortable.
You can query S3 like a database. With S3 Select and Athena, you can run SQL directly against CSVs and JSON files sitting in a bucket. No cluster. No database engine. Pay only when you query. For analytics workloads at startup scale, this changes everything.
The presigned URL is one of the most elegant security patterns in AWS. Your server signs a temporary URL. The browser uploads directly to S3. Zero bandwidth cost on your server, zero exposed credentials, zero public bucket. Five lines of code.
Since posting it on the AWS Builder Center, the response has genuinely surprised me.
The comment thread has been active — builders sharing their own "wait, I never knew that" moments with S3. Several people mentioned they'd been using AWS for years and had never questioned the folder metaphor. A few architects chimed in about the data lakehouse pattern and how it's changed their infrastructure decisions. One comment that stuck with me said something like: "This is the kind of article that reminds you to slow down and actually learn your tools."
That's exactly what I was going for.
If your post is getting traction, add your live stats here — upvotes, comments, shares 🙂
This got me thinking: how many other AWS services do we use daily on autopilot without ever going deeper?
I'm planning a follow-up piece — probably on AWS Lambda's execution environment and the cold start myth that most people misunderstand. Same idea: familiar service, surprising depth.
If that sounds interesting, follow along here or connect with me on the AWS Builder Center.
Have thoughts on S3 or anything I got wrong? Drop a comment below — I'd love the discussion.
Tags: #AWS #S3 #CloudComputing #AWSBuilder #Serverless #TechBlog #CloudArchitecture
