Hello there. New to meteor and first thread. 
I’ve been trying to read up on using Meteor at scale. Most of the discussions are generic and higher level, proxies and multiple instances 101.
One glaring omission I noticed is the lack of any mention of object pooling. When running an app on a VM under load pooling is downright necessary to tame it’s garbage collector. Apps tend to be a tad less responsive when you have long duration garbage spikes!
I see from google that the poor node.js folks have figured this out the hard way, but haven’t yet found any Meteor articles even mentioning it.
Is there something about meteor that precludes this that I’m unaware of? Or is it just that the nature of meteor somehow doesn’t tend to involve a high frequency of object creation?