Version: 1.12.0 (using 4.2.96 (KDE 4.2.96 (KDE 4.3 RC2)) "release 142", KDE:KDE4:Factory:Desktop / openSUSE_11.1) Compiler: gcc OS: Linux (x86_64) release 2.6.27.23-0.1-default Randomnly kontact starts using a lot of cpu usually more than 50%. I did an ltrace on the process, I hope may be of some help. Alin
Created attachment 35537 [details] ltrace log for kontact high cpu load
Please give a list of which resource types you have registered in Contact and Calendar modules, and which account types you use in kmail.
Hi, kmail are imap accounts 2 google 2 other resources: *Birthdays *2 calendar in remote file resources *1 akonadi resource ** with some previous imported resources when migration to akonadi was done ** 3 resources that were added to akonadi all of them are calendars in remote files Contact Default address book Alin
got rid of the akonadi_resource from kontact attached with gdb here is the bt (gdb) bt #0 0x00007fd5da24b386 in poll () from /lib64/libc.so.6 #1 0x00007fd5d59ac768 in ?? () from /usr/lib64/libglib-2.0.so.0 #2 0x00007fd5d59aca8b in g_main_context_iteration () from /usr/lib64/libglib-2.0.so.0 #3 0x00007fd5dc0fdd3f in QEventDispatcherGlib::processEvents (this=0x60d870, flags=<value optimized out>) at kernel/qeventdispatcher_glib.cpp:327 #4 0x00007fd5dae89fef in QGuiEventDispatcherGlib::processEvents (this=0x266dfb0, flags=<value optimized out>) at kernel/qguieventdispatcher_glib.cpp:202 #5 0x00007fd5dc0d31d2 in QEventLoop::processEvents (this=<value optimized out>, flags={i = -391892992}) at kernel/qeventloop.cpp:149 #6 0x00007fd5dc0d35a4 in QEventLoop::exec (this=0x7fffe8a43040, flags={i = -391892912}) at kernel/qeventloop.cpp:201 #7 0x00007fd5dc0d5894 in QCoreApplication::exec () at kernel/qcoreapplication.cpp:888 #8 0x0000000000404995 in _start ()
That's just the even toop. Try disabling resources methodically to isolate the problem to one resource. Also try running kmail and korganizer and kaddressbook standalone.
So? can we get feedback about this report please.
Hello, I can confirm this bug in karmic using kde 4.3 fromm ppa. I have only one account configured as Disconnected IMAP of a kolab server. The problem is actually with korganizer because if I kill it the CPU goes down to standard usage. Korganizer is launched like this by kontact: /usr/bin/korgac -icon korgac Futhremore closing kontact does not close korgnaizer so once you launched kontact once, you have to kill KOrgnaizer to be able to use your machine correctly.
I suspect this is gone in 4.4.0+
Changing the bug status as discussed on irc.
Regression in KDE PIM 4.4.5 with akonadi and if a mail-resource is added to akonadi. In this case, the kontact process consumes 98% of available I/O bandwidth thus slowing the whole system to near unresponsiveness. Blame nepomuk and its unefficiencies. But the process being visible doing the I/O in iotop is kontact.