eventmachine and rack in particular are pure platform things. you have to be really careful, so many rubies and OSes and clients
@raggi perhaps people are starting to realize that a lot of apps max out at 100-1000ish connections and threads are FINE
-
-
@bascule sure, but we do have production users of EM who balance to ~4500 concurrent users under nominal load, spiking to 10k -
@raggi EventMachine definitely roflscales better ;) People still use it to juggle a handful of connections -
@bascule it'd be a better use of my time to make celluloid roflscale than to fix EM. sadly, Rack takes priority with what time i have -
@raggi please fix Rack ;) -
@bascule confirm, asap
End of conversation
New conversation -
-
-
@bascule it really depends what you're building. EM is great for things that essentially equate to distributors or DPI proxiesThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@bascule basically the things that ZMQ is good at, but on raw TCP (functionally). ZMQ has much more power elsewhereThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@bascule also, i don't know what your throughput metrics look like these days, how easily can you fill 100MB/s? -
-
-
@raggi pretty sure Wildfire pushes a lot more data than anyone is pushing through Celluloid ;) -
@bascule dude, we barely use it. oh, except when we hit the top of the Facebook API dashboard for a week solid… yeah, that was a fun call
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.