Trying to figure that out now. DB seems spiffy, and express is logging requests just fine. Thats the trivial stuff - now I need to start getting my hands dirty.

EDIT - OK serious issues now. getting a range error on redis and the call stack is blowing out, recycling node every 15 seconds or so. Will let you know when its back up.

EDIT2 - Turned out to be an exception due to an invalid date string. Fixed!

Along the same lines, what are some of the metrics of your traffic that caused socket.io to choke (total connections, # concurrent, etc)?

or was it the other pieces in your stack that caused the bottleneck.

BTW, great idea.