You containerized a Lambda function. Then you moved it to AWS Fargate, and it just...worked.

But here's the part I glossed over: Fargate isn't a service you deployed to. It's a launch type. When you ran that task, you had your first encounter with Amazon ECS — and you quietly accepted a bunch of defaults you didn't even know you were choosing.

So what is ECS, really?

You might be considering whether to deploy your application to a serverless service. Maybe you're an AWS customer weighing AWS Lambda (serverless functions) against AWS Fargate for Amazon ECS (serverless containers).

Your application fits within Lambda's hard constraints — 15-minute max duration, package size, payload limits, memory, and cold start latency all check out.

Is Lambda a good fit? Is it better to run on AWS Fargate?

It's an important question, especially in organizations that deploy a lot of resources to AWS. In this post, we'll work through that question with a focus on cost — and we'll cover a technique you can use to minimize the risk of getting the answer wrong.

If you've been following this series, you know what AWS Lambda Managed Instances (LMI) are and what they cost. In Part 1 we covered the overview and in Part 2 we dug into the cost model. Now it's time to answer the question that actually matters for day-to-day development: will my existing Lambda code work on LMI, and what will I need to change?

The short answer is: maybe - and possibly quite a bit.

The execution model is fundamentally different from on-demand Lambda. Don't assume your code will behave the same way - some of these differences will silently corrupt your data, and others will just make your life harder. Read through all of them before migrating anything to production.

You've likely seen the AWS launch post and are intrigued by what AWS Lambda Managed Instances (AWS LMI) can offer. You may have even read my overview of AWS LMI in an earlier blog post.

In this post, we'll look at AWS Lambda On-Demand costs, how costs change with AWS LMI, when it makes sense to use AWS LMI and when it makes sense to stick with On-Demand AWS Lambda.

You've optimized your Lambda functions six ways from Sunday, but costs are still eating your budget. AWS LMI might help.

You've bought in to using containers to package AWS Lambda functions but now you're curious...

If you had to, could you migrate that same function to AWS Fargate?

You're a Docker pro - or maybe you just work somewhere with strong opinions about how containers should be built. Either way, let's talk about bringing your own container images to AWS Lambda.

You're a Go (or Rust) developer and you're really looking forward to deploying your awesome Lambda application in a OS-only container. Let's look at how to do it.

Your Lambda function just hit the 250MB unzipped size limit.

Again.

You're wrestling with Layers and code dependencies, wondering if there's a better way.

There is — and it might surprise you how simple container packaging makes this problem disappear.

Welcome to the Curious Dev Blog! This is where I'll explore the fascinating world of software development, share insights, and dive deep into the technologies that shape our digital world.