On Feb 10, 2010, at 2:05 PM, Emmanuel Lecharny wrote:
On 2/10/10 10:28 PM, Alan D. Cabrera wrote:
each of those tasks can be processed by a separate thread selected
in a thread pool. IMO, it may be better than the current
architecture where we have a pool of IoProcessor, each one of them
having its own Selector, and no thread to process the events. For
instance, if we have 3 IoProcessor (the default value for a dual
core processor), then we can only process 3 events in parallel.
Pretty inefficient...
I tried to follow the code for IoProcessor and my brain hurts. :)
I share the same feeling ...
Did I read correctly that the sessions are partitioned between N
IoProcessors?
Absolutely. Round robin.
Then I definitely like what you're proposing. Things are more up
front and the concurrency is more visible, imho.
Regards,
Alan
