Modeling and Analyzing Lambda vs. Fargate Breakeven

I enjoyed reading this analysis, even if it is quite specific to a single situation. I agree with the writer's larger point that you need to do this kind of analysis for your workload, in your conditions, to draw the right conclusions.

> This does lead to a caveat with analysis, however: larger Lambda’s may actual process more quickly than ones that are more right sized, potentially leading to cheaper outcomes.

This is an important thing to include for compute-bound tasks (which is a suprisingly large number of tasks, at least up to some limit). There are some nice third party projects (like Lambda Power Tuning[2]) that can help find the sweet spot on price versus performance (with their own caveats, of course).

> Consider using Lambda without API Gateway

Yes! API Gateway can add a lot of value when you need it, but if you don't need it there's no reason to carry the cost. Other options are to use the SDK (as the article says), Lambda Function URLs[1], or Lambda with ALB.

[1] https://aws.amazon.com/blogs/aws/announcing-aws-lambda-funct... [2] https://github.com/alexcasalboni/aws-lambda-power-tuning

Really appreciate this! great food for thought and highlights the answer is often ... "it depends!" :-)

Lambda and Fargate are both container services.

The real point here, IIUC, is whether having the container run continuously makes sense.

There are many other directions to take the discussion, e.g. K8s, but constraining it enough to compare the two options makes this an excellent article.

From 2019 but certainly still relavent