Fun fact: sub 5mb uncompressed payload for a Lambda equals around 200-300ms coldstarts. One *can* upload 250mb+ payloads if you want to hack around it and have SUPER SLOW coldstarts. https://hackernoon.com/exploring-the-aws-lambda-deployment-limits-9a8384b0bec3 … But you don't want to do that. More and smaller lambdas is faster.
-
-
Yea it works but not enough to warrant it in our codebase. I'll probably add to arc for prod lambdas eventually.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Straight up Prune from TJ which just gets rid of non-js files makes a big difference
-
Ah I had forgotten about that one!
-
I have no data to back this up, but feels like the bigger win of bundling may be eliminating all the require calls and file loading.
-
In our app it shaved around 80-100ms which def helps but those Lambdas are smol to begin with; imagine a huge payload 10+mb it makes a big dent. That said: just don't do that (if you can).
-
I keep saying this (and not doing it) but would be really nice to have some standard benchmarks for this stuff. Something like TodoMVC that could be run across providers.
-
TL;DR:
AWS
.
.
.
. ~seconds elapse~
.
.
.
Everyone else - End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
