Emmanuel, turns out that the code I've previously included did not test my issue in production exactly. The issue here was due to utilizing Executors.newFixedThreadPool(threadCount) in the SocketAcceptor, and when I switched over to Executors.newCachedThreadPool(), I did not see any pauses. The Mina documentation recommends using newCachedThreadPool for socketacceptor as it may lead to weird performance issues which I've faced recently. The notable fact here is that on a 32bit system Redhat 5.1 running JDK1.6, Executors.newFixedThreadPool performed flawlessly, and that's why I've been using it for last 2 years. But when we moved up to 64Bit CentOS the newFixedThreadPool produced weird results which originated this whole thread.
Thanks, Frank On Wed, Apr 21, 2010 at 2:37 AM, Emmanuel Lécharny <[email protected]>wrote: > >> Emmanuel, thanks for the help, I think I narrowed it down to socket >> acceptor >> thread pool being stuck on "waiting on condition." >> >> > Can you test your code with 1.1.7 ? > > > -- > Regards, > Cordialement, > Emmanuel Lécharny > www.nextury.com > > >
