Can you explain in simple words what does it means?
I can see in many sites that they recommend using lambda for saving , and now you say that when you get lucky and grow the bill will explode?
Keeping a server running costs you X/hour. Having a lambda costs you Y/invocation. Which is cheaper depends on the number of invocations per hour. If you have 1 invocation per hour, a lambda will be cheaper. If you have 2 million per hour, a server might be cheaper.
Eventually, they get lucky, and their Lambda functions are running non-stop, incurring a fairly large premium over a normal constantly running VM.
I managed to do this with my side project (AWS bill in the several hundred dollars per month range), I wrote about it here: https://onlineornot.com/on-moving-million-uptime-checks-onto...