Shared memory with Node.js
blog.varunajayasiri.com
blog.varunajayasiri.com
Node is great for squeezing performance out of a box saddled with io bound tasks, or handling a large number of mostly discrete connections, that don't share a ton of state.
If you need access to something like a cache where synchronization is not a major issue, and you don't want to dupe memory across the processes, you can throw something like redis on the box.
However, if what you really want shared memory across what are now essentially threads...why wouldn't you go to a technology that's been designed for the task? I'm referring to the likes of Java, Scala, golang, etc.
[1] http://swannodette.github.io/mori/
The "tricky party" is when concurrent processed need to synchronize some reference. Facilities for such ops can be orthogonal to the implementation of immutable collections, and then the two can be used together to good effect.
http://hypirion.com/musings/understanding-persistent-vector-...