[Bug 691105] New: System freezes randomly and repeatedly under heavy load
https://bugzilla.novell.com/show_bug.cgi?id=691105 https://bugzilla.novell.com/show_bug.cgi?id=691105#c0 Summary: System freezes randomly and repeatedly under heavy load Classification: openSUSE Product: openSUSE 11.4 Version: Factory Platform: x86-64 OS/Version: openSUSE 11.4 Status: NEW Severity: Critical Priority: P5 - None Component: Other AssignedTo: bnc-team-screening@forge.provo.novell.com ReportedBy: pushkin@email.cz QAContact: qa@suse.de Found By: --- Blocker: --- User-Agent: Opera/9.80 (X11; Linux x86_64; U; cs) Presto/2.8.131 Version/11.10 When under heavy load (e.g. Finite Element Method calculations) the system freezes. First hangs X and KDE (mouse cursor moves), then if switched to console it is possible to run some commands (except kill - this freezes chosen console), then the system freezes completely. init 0, init 6 and shutdown -h now lead also to system freeze. The bug seems to be very similar to this one: https://bugzilla.novell.com/show_bug.cgi?id=684634 but I have no hardware nVidia components. Reproducible: Always Steps to Reproduce: 1. Run some application for obtaining high CPU consumption 2. Do some usual work 3. After a few hour the system freezes Actual Results: The system freezes. Expected Results: Normal work. Happens at two different hardware configurations, first: CPU: AMD Phenom(tm) II X4 945 Graphics: ATI RadeonHD 4650 Kernel: 2.6.37.1-1.2-desktop Driver: ATI Catalyst 11.3, installed manually second configuration: APU: AMD E-350 Processor Kernel: 2.6.37.1-1.2-desktop Driver: ATI Catalyst 11.2, installed from ATI Repository The same happened with kernel 2.6.37.6-18.1-desktop from BuildService repository: http://download.opensuse.org/repositories/Kernel:/openSUSE-11.4/openSUSE_11.... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c
Karel Hruška
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c1
--- Comment #1 from Karel Hruška
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c
zj jia
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c2
Michal Hocko
Are you able to reproduce without X server running? Is this memory or CPU/scheduler related? (E.g. are you able to reproduce just with a simple cpu hog like busy loop). I cannot perform these tests at the moment, I will try them as soon as
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c3
--- Comment #3 from Karel Hruška
How many processes are running/existing in the system when the hang appears? I am not sure, I hope it should not be too many processes ... cca 200? I will watch for that next freeze.
Is there any free memory (do you have swap enabled)? Yes, memory usage is at circa one half, swap is enabled with 100G space, it is almost empty all the time.
Could you run some diagnostic in the background. vmstat 1 cat /proc/schedstat (in a loop once in a while) cat /proc/sched_debug (when the problem appears) Of course, I have made a script logging vmstat and schedstat. I will attach results after next failure.
In the meantime I have switched from KDE to Fluxbox and run my computations again. I will let you know about results. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c4
--- Comment #4 from Karel Hruška
In the meantime I have switched from KDE to Fluxbox and run my computations again. I will let you know about results.
Currently I am running over two days without any freeze in Fluxbox, therefore I hope that the problem with freezes is a KDE-related issue. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c5
--- Comment #5 from Karel Hruška
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c6
--- Comment #6 from Karel Hruška
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c7
--- Comment #7 from Karel Hruška
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c8
--- Comment #8 from Karel Hruška
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c9
--- Comment #9 from Karel Hruška
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c10
--- Comment #10 from Karel Hruška
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c11
--- Comment #11 from Karel Hruška
Are you able to reproduce without X server running? Is this memory or CPU/scheduler related? (E.g. are you able to reproduce just with a simple cpu hog like busy loop). How many processes are running/existing in the system when the hang appears? Is there any free memory (do you have swap enabled)?
Could you run some diagnostic in the background. vmstat 1 cat /proc/schedstat (in a loop once in a while) cat /proc/sched_debug (when the problem appears)
-- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c12
Stefan Dirsch
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c13
Ursan Marius Bogdan
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c
Jiri Slaby
https://bugzilla.novell.com/show_bug.cgi?id=691105
https://bugzilla.novell.com/show_bug.cgi?id=691105#c14
Jeff Mahoney
participants (1)
-
bugzilla_noreply@novell.com