I think one of the main things to be aware of, with serverless, is that it is NOT as simple as it’s made out to be. Sure, you can get a lambda (talking AWS here specifically) function up and running easily enough and you go “Yay, so easy”, but after a while you start to realise that there are almost innumerable “gotchas” that are poorly documented and very seldom mentioned by those evangelising the process.
Don’t get me wrong, I’m fully committed to serverless. It’s the way of the future and it’s a future I’ve been anticipating for 20+ years. But don’t underestimate the complexity that you’re signing up for.
I’ll probably write an article, but the number of cases I’ve encountered that required a “Oh, crap” rethink based on fundamental limitations on deployment, configuration, event planning, pricing calculations and all that jazz will probably need a book.
All the best from someone who’s building a 100% serverless, sophisticated application.
And not missing servers!