> My company ran into this issue using lambdas to process some data where the 99% of requests were fine running in 256mb but a few required more. Well, you co

> My company ran into this issue using lambdas to process some data where the 99% of requests were fine running in 256mb but a few required more. Well, you could have deployed 2 lambdas: one as a coordinator with 256MB and another worker with the Ram you need called in those 1% of cases. And remember that if you had to do that on a server with a virtual machine you still needed all the ram deployed 24/7.