2006/4/20, Gustavo Noronha <[EMAIL PROTECTED]>: > > Like I said in a previous email, I know about dumping thread information, > but I am unnable to identify what thread is using up the CPU cycles. What I > did to try to fix this problem is setting up some different c3p0 > configurations: > > on hibernate.properties: > > hibernate.c3p0.idle_test_period=200 > hibernate.c3p0.acquire_increment=3 > > on c3p0.properties (newly created): > > c3p0.numHelperThreads 8 > > I haven't seen c3p0 problems up to now, nor 100%-cpu-using processes > simply segfaulting and futexing, but it's still early to decide whether it > fixed the problem or not. I'd still like to have insights on where the > problem may lie. I am afraid to be work-arounding the problem with a c3p0 > configuration when in fact the application is misbehaving with the > connections, for instance. >
So, just to let you know, after this configuration was done I haven't seen the problems I mentioned yet. The one other thing I did was increasing the maximum size of the jvm heap, after seing an OutOfMemory error, but these errors only happened once, and do not seem to be related to the other problem. The system was running fine days before I did that change, too. Thanks for all the help! -- Gustavo Noronha Silva Coordenação de Segurança e Sustentação Ministério do Desenvolvimento Social - Brasil
