Large Applications on AWS Lambda

Zappa has an auto-keep warm features, so cold-boot is largely negated.

Also, this is a feature really designed for scientific computing, which generally happens asynchronously anyway, versus say something like web-app serving.

Really the big drawback here is that it sucks up available RAM, so you may still have to do some tweaking on a per-application basis.

Still, this feature was contributed out of a user's own necessity, so I'm certain that others will get use out of it too.