Application: akonadi_nepomuk_contact_feeder (0.1) KDE Platform Version: 4.4.4 (KDE 4.4.4) Qt Version: 4.6.2 Operating System: Linux 2.6.32-22-generic x86_64 Distribution: Ubuntu 10.04 LTS -- Information about the crash: Using Kubuntu 10.04 with ppa enabled (KDE 4.4.4). On reboot akonadi tried starting up and crashed. -- Backtrace: Application: Akonadi Agent (akonadi_nepomuk_contact_feeder), signal: Segmentation fault [KCrash Handler] #5 0x00007fe41cbc6f60 in KMimeType::mimeType (_name=..., options=KMimeType::ResolveAliases) at ../../kdecore/services/kmimetype.cpp:159 #6 0x00007fe41c7daa22 in Akonadi::MonitorPrivate::isMimeTypeMonitored (this=0xf22d70, msg=...) at ../../akonadi/monitor_p.h:200 #7 Akonadi::MonitorPrivate::isItemMonitored (this=0xf22d70, msg=...) at ../../akonadi/monitor_p.h:82 #8 Akonadi::MonitorPrivate::acceptNotification (this=0xf22d70, msg=...) at ../../akonadi/monitor_p.cpp:141 #9 0x00007fe41c7db578 in Akonadi::MonitorPrivate::slotNotify (this=0xf22d70, msgs=<value optimized out>) at ../../akonadi/monitor_p.cpp:287 #10 0x00007fe41c754102 in Akonadi::ChangeRecorderPrivate::slotNotify(QList<Akonadi::NotificationMessage> const&) () from /usr/lib/libakonadi-kde.so.4 #11 0x00007fe41c7d6a3a in Akonadi::Monitor::qt_metacall (this=0xf57200, _c=QMetaObject::InvokeMetaMethod, _id=<value optimized out>, _a=0x7ffffd63f400) at ./monitor.moc:147 #12 0x00007fe41c7512b8 in Akonadi::ChangeRecorder::qt_metacall (this=0x7ffffd63f170, _c=377592560, _id=-43781664, _a=0x1) at ./changerecorder.moc:72 #13 0x00007fe41c3c5e3f in QMetaObject::activate (sender=0xf57220, m=<value optimized out>, local_signal_index=<value optimized out>, argv=0x1) at kernel/qobject.cpp:3293 #14 0x00007fe41c821602 in OrgFreedesktopAkonadiNotificationManagerInterface::notify (this=0x7ffffd63f170, _t1=<value optimized out>) at notificationmanagerinterface.moc:85 #15 0x00007fe41c82164d in OrgFreedesktopAkonadiNotificationManagerInterface::qt_metacall (this=0xf57220, _c=QMetaObject::InvokeMetaMethod, _id=<value optimized out>, _a=0x7ffffd63f670) at notificationmanagerinterface.moc:73 #16 0x00007fe41bdd3486 in QDBusConnectionPrivate::deliverCall (this=<value optimized out>, object=<value optimized out>, msg=<value optimized out>, metaTypes=..., slotIdx=<value optimized out>) at qdbusintegrator.cpp:904 #17 0x00007fe41bdde05f in QDBusCallDeliveryEvent::placeMetaCall(QObject*) () from /usr/lib/libQtDBus.so.4 #18 0x00007fe41c3c2d49 in QObject::event (this=0xf57220, e=0x131f8f0) at kernel/qobject.cpp:1248 #19 0x00007fe41a0bd22c in QApplicationPrivate::notify_helper (this=0xdf3640, receiver=0xf57220, e=0x131f8f0) at kernel/qapplication.cpp:4300 #20 0x00007fe41a0c36fb in QApplication::notify (this=0x7ffffd640130, receiver=0xf57220, e=0x131f8f0) at kernel/qapplication.cpp:4183 #21 0x00007fe41d15c606 in KApplication::notify (this=0x7ffffd640130, receiver=0xf57220, event=0x131f8f0) at ../../kdeui/kernel/kapplication.cpp:302 #22 0x00007fe41c3b306c in QCoreApplication::notifyInternal (this=0x7ffffd640130, receiver=0xf57220, event=0x131f8f0) at kernel/qcoreapplication.cpp:704 #23 0x00007fe41c3b57e7 in QCoreApplication::sendEvent (receiver=0x0, event_type=<value optimized out>, data=0xddcc60) at ../../include/QtCore/../../src/corelib/kernel/qcoreapplication.h:215 #24 QCoreApplicationPrivate::sendPostedEvents (receiver=0x0, event_type=<value optimized out>, data=0xddcc60) at kernel/qcoreapplication.cpp:1345 #25 0x00007fe41c3dc9d3 in QCoreApplication::sendPostedEvents (s=<value optimized out>) at ../../include/QtCore/../../src/corelib/kernel/qcoreapplication.h:220 #26 postEventSourceDispatch (s=<value optimized out>) at kernel/qeventdispatcher_glib.cpp:276 #27 0x00007fe41685a8c2 in g_main_context_dispatch () from /lib/libglib-2.0.so.0 #28 0x00007fe41685e748 in ?? () from /lib/libglib-2.0.so.0 #29 0x00007fe41685e8fc in g_main_context_iteration () from /lib/libglib-2.0.so.0 #30 0x00007fe41c3dc513 in QEventDispatcherGlib::processEvents (this=0xddc3b0, flags=<value optimized out>) at kernel/qeventdispatcher_glib.cpp:412 #31 0x00007fe41a16d46e in QGuiEventDispatcherGlib::processEvents (this=0x7ffffd63f170, flags=<value optimized out>) at kernel/qguieventdispatcher_glib.cpp:204 #32 0x00007fe41c3b1992 in QEventLoop::processEvents (this=<value optimized out>, flags=) at kernel/qeventloop.cpp:149 #33 0x00007fe41c3b1d6c in QEventLoop::exec (this=0x7ffffd6400b0, flags=) at kernel/qeventloop.cpp:201 #34 0x00007fe41c3b5aab in QCoreApplication::exec () at kernel/qcoreapplication.cpp:981 #35 0x00007fe41c7356a6 in Akonadi::AgentBase::init (r=0xf56500) at ../../akonadi/agentbase.cpp:496 #36 0x0000000000408dd8 in _start () Possible duplicates by query: bug 236600. Reported using DrKonqi
*** Bug 241237 has been marked as a duplicate of this bug. ***
*** Bug 241748 has been marked as a duplicate of this bug. ***
*** Bug 241816 has been marked as a duplicate of this bug. ***
I have also noticed something interesting. When you start akonadi at the same time you start nepomuk you get a crash on the nepomukstorage service. This is very consistent and could be the cause for this crash. Just shut down both akonadi and nepomuk and then, from a console do: nepomukserver && akonadictl start That's it, you get an unusable nepomuk as a result.
*** Bug 240482 has been marked as a duplicate of this bug. ***
*** Bug 242621 has been marked as a duplicate of this bug. ***
*** Bug 243288 has been marked as a duplicate of this bug. ***
Hej, we have received reports about this bug only from Ubuntu users so far, so I really guess it's an (K)Ubuntu specific issue... Ciao, Tobias
Tobias, I'm running openSuse 11.2, and I'm getting the same crash on every login (KDE 4.4.4 and 4.4.5). This is a very annoying bug.
*** Bug 245544 has been marked as a duplicate of this bug. ***
*** Bug 246046 has been marked as a duplicate of this bug. ***
*** Bug 246441 has been marked as a duplicate of this bug. ***
Hej again, can all of you, who can reproduce this crash, please execute kbuildsycoca4 --noincremental and test whether that helps? Ciao, Tobias
On Tuesday, August 03, 2010 02:55:01 am Tobias Koenig wrote: > https://bugs.kde.org/show_bug.cgi?id=240981 > > > Tobias Koenig <tokoe@kde.org> changed: > > What |Removed |Added > --------------------------------------------------------------------------- > - Status|UNCONFIRMED |NEEDSINFO > Resolution| |WAITINGFORINFO > > > > > --- Comment #13 from Tobias Koenig <tokoe kde org> 2010-08-03 09:55:00 --- > Hej again, > > can all of you, who can reproduce this crash, please execute > kbuildsycoca4 --noincremental > and test whether that helps? > > Ciao, > Tobias Tobias, I am new to this and don't know if I am replying in the right place. I did as you asked and did not notice that it made any difference. Although Akonadi is not crashing as frequently as it was when I reported this bug. Chris
(In reply to comment #13) > can all of you, who can reproduce this crash, please execute > kbuildsycoca4 --noincremental > and test whether that helps? Does not seem to help here.
*** Bug 242005 has been marked as a duplicate of this bug. ***
*** Bug 242251 has been marked as a duplicate of this bug. ***
Since kbuildsycoca4 --incremental doesn't fix the bug, I reopen it. Unfortunately none of the developers here can reproduce this bug, which makes fixing it impossible :/
And what about distribution specific (Ubuntu), architecture (64bit) or auto login feature? I have Ubuntu 10.4, 64bit and auto login enable
*** Bug 246967 has been marked as a duplicate of this bug. ***
Since I upgraded to the 4.5 RC packages I was not able to reproduce this bug. My gut feeling is this has to do with a race condition at startup, as even then I couldn't reproduce it only at logon. If anyone was able to reproduce the bug by recreating the race condition I'd be happy to help more. Just tell me what commands to run. Good hunting, Tsvi
*** Bug 247531 has been marked as a duplicate of this bug. ***
This Bug is reproduceable here, Ubuntu Lucid 10.04.1 + PPA KDE 4.4.4 as mentioned before. We found the problem is especially occouring on coperate setups with $HOME ode $USER on NFS Servers. If an user logs on using an other worstation, which is using $HOME of $USER the mysql embedded backend crashes/destroys the data. Also it is impossible to use the Resources on any other workstaions. In other words: critical failure. In support cases we recommend to use thunderbird als running email client within those call ins. KDE PIM is critically broken - unusable.
(In reply to comment #23) Hej Jürgen, > We found the problem is especially occouring on coperate setups with $HOME ode > $USER on NFS Servers. > > If an user logs on using an other worstation, which is using $HOME of $USER the > mysql embedded backend crashes/destroys the data. You mean if the user has running _two_ sessions in parallel from different workspaces? Well, that will indeed will cause troubles, but this will happen with any other software as well, which assumes that it has exclusive access to its configuration/data files. Ciao, Tobias
> You mean if the user has running _two_ sessions in parallel from different > workspaces? Well, that will indeed will cause troubles, but this will happen > with any other software as well, which assumes that it has exclusive access to > its configuration/data files. Yes, this is the Problem. Akonadi is not protected against this. So this is an critcal "NOGO-AT-ALL ERROR! It is an very bad, bad idea and short thought concept, that this an problem at all. KDE is definitely not single an USER GUI! Multiple Logons are definetly an "Must work" feature. Especially in Enterprise environments. Embedded MySQL Datafiles is an bad idea - IMHO a broken concept - even upon this issue. "Works here local" points of arguements are definitely not an Option at all. So this Bug has absolute Priority. Jürgen Sauer
Or on the other hand Akonadi and KDEPIM stays as it is now, an unusable toy with build in data looses. J. Sauer
So any idea how to fix it? Bashing around is easy, but constructive critic is more helpful!
The original crash report seems to be fixed in KDE 4.5.1 At least I don't run into it anymore. Also, there seems to be no more races between akonadi and nepomuk.
(In reply to comment #25) > > You mean if the user has running _two_ sessions in parallel from different > > workspaces? Well, that will indeed will cause troubles, but this will happen > > with any other software as well, which assumes that it has exclusive access to > > its configuration/data files. > Well, it seems to me that your problem is concurrent access to the database files. From my point of view there are two solutions to your problem: 1) On your side:Configure akonadi to connect to a dedicated mysql server with a database for each user of your system. Then, concurrent accesses will be dealt with by the server with no problems. 2) On development side: Get akonadi to recognize that the database is already being accessed by another machine and take steps to not colide with it. Ideally, akonadi would make a copy of the database files and operate on those in the case of a concurrent access. Concurrent access to resources seems to already be handled properly. I hope this helps, Luis
*** Bug 256289 has been marked as a duplicate of this bug. ***
*** Bug 252228 has been marked as a duplicate of this bug. ***
Created attachment 56302 [details] New crash information added by DrKonqi still occures with KDE 4.4.5 directly after boot. At last shutdown the debugging windows where showing directly after logoff (they were kind like in behind my X-Server-Session)
Idea for possible best practice: On first Longon/first use on akonadi an user should be asked for place to store the pim data. As possible options should be: - postgres server @ an dedicated server to avoid datalooses on nfs shares, and licence proplems using mysql. (MySQL is not LGPL! thats an huge not-possible-at-all problem!) - where to store local datafiles - an makes clear information flow -> user for such problems. On the other hand the akonadi developers should test the code before releasing. A complete test musst include Multiuser environments, NFS Setups, Multiple logins etc. The problem was raising due bad and incomplete tests, and short-thinking Bye
*** Bug 264466 has been marked as a duplicate of this bug. ***
from bug 272395 I tried to overwrite the std.ics file of KCalendar while the application was open. Then Akonadi crashed. I had several other windows opened, but I believe they were mostly inactive. It is possible that some copying was still in place. Thanks!
*** Bug 272395 has been marked as a duplicate of this bug. ***
*** Bug 277563 has been marked as a duplicate of this bug. ***
*** Bug 275205 has been marked as a duplicate of this bug. ***
*** Bug 268502 has been marked as a duplicate of this bug. ***
*** Bug 274136 has been marked as a duplicate of this bug. ***
*** Bug 280398 has been marked as a duplicate of this bug. ***
Since I have installed ONEIRIC OZELOT, Akonadi and his subprocesses Nepomuk, cid, Kmail etc crash on startup. So I can't execute some applications (like openbve f.ex.) and the computerspeed goes down. When I use the ksysguard it shoes me that nepomuk and cid and knotify use even 45% of CPU (when they are not crashed) On the system before (natty) all worked great but I think they changed the paths to the logs and so nepomuk etc can't work any more
Unsure if my problem is related, but I post here: After recent upgrade to the 3.1 kernel (and some other packages, don't remember which), my computer freezes completely about 1 min after login. I installed a different window manager and the problem went away.. Until I started Kontact. Then, it took about 30 sec for the freeze to appear. I am hence GUESSING this is related to Akonadi. The bug for me appears every time, and the only choice I've found is to force a shutdown of the computer. I have waited several minutes on several occations to see if it awakens again, but no luck. I have also tried to remove .config/akonadi, removed all config files I could find in .kde4, and .local/share/akonadi*. Then, I see the server self-test appear with some errors before it freezes. Since that didn't work, I renamed the entire .kde4, but it still froze. I will try the "kbuildsycoca4 --incremental" afterwards, right now I am checking one final time if it reacts after a while. I am using Arch Linux, 64bit, with latest KDE 4.7.3. Interestingly enough I have the exact same setup on my laptop and desktop, but the problems only occur on my laptop. I am inclined to believe I have screwed up some settings. I am also inclined to believe the my bug is unrelated to what is reported here, as I have not had any issues in the past. Suggestions are highly welcome!
Update: kbuildsycoca4 did not help, the computer still freezes.
New information: I am quite lost as to understand where I should report this problem. As I said, my desktop does not have this problem at all, while for my laptop the problem occurs every time. Now, I have discovered however that I can trigger the bug on my desktop by running "git svn rebase" on a git mirror I have of a svn repository. I have tried this twice, once I left it running while I ate dinner just to check. I also noticed that I cannot log into my desktop using ssh, so I guess the entire computer has pretty much locked up. My best assumption so far is that it is related to the recent upgrade to 3.1, but since I don't seem to find similar bug reports elsewhere, I am unsure what to think. ANY suggestions on what to look for in /var/log, or what I could try to do in order to figure out what is going on is highly welcome!
*** Bug 295203 has been marked as a duplicate of this bug. ***
There haven't been any new reported crashes for this bug for over 1.5 years now. Given that the code base of the feeder has changed substantially over the time, I'm marking this bug as FIXED.