If the server you're targeting is quite slow in terms of sending a response back, let's say its 500ms on average, isn't this an expensive exercise? I mean, AWS Lambda charges for every 100ms of utilisation.
Amazon charges 1/50,000th of a penny per 100ms, so 1 second is still 1/5000th of a penny. You need 5000 seconds so well over an hour (90 minutes) to pay Amazon a measly 1 cents for an hour of their resources.
