The comments posted on the original article are not impressed.

> One database query took longer than expected. That blocked an async task. Which held up the executor.

This only happens if you're using blocking code in the event loop, right? This article seems like an admission of over-engineering from the get go.

The first thing I thought when I read this weird article was "wait...do they use their database synchronously in an asynchronous environment or am I missing something here?".

Most of us have been bitten by this naive mistake and I'm sure we have learned our lesson.