You’re preaching to the choir :)

The point of this reproduction was to do the opposite of the AWS recommendations to show what happens, and explain why it happens (most of that is in my first post, though). Once one understands that, they should understand how the mitigations help and why they’re only mitigations not fixes.

Low concurrency limits with SQS queues are playing with fire, so you need to take the right precautions.

I’m writing from a place of experience as a few teams at work have ran into this issue. It’s not an intuitive problem and most people don’t read the AWS docs in depth. Even less read my blog posts, but that’s not that point, haha.