Comment by mactunes

How do you ensure that the 500 concurrent keep-alives land on different Lambda functions? I.e. requests 220 and onwards might hit lambda functions which were warmed up by requests 0-219. I just made the above numbers up of course.